1
0
mirror of https://github.com/open-telemetry/opentelemetry-go.git synced 2024-12-12 10:04:29 +02:00
opentelemetry-go/sdk/trace/span.go

597 lines
16 KiB
Go
Raw Normal View History

// Copyright The OpenTelemetry Authors
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package trace // import "go.opentelemetry.io/otel/sdk/trace"
import (
"context"
"fmt"
"reflect"
"sync"
"sync/atomic"
"time"
"go.opentelemetry.io/otel/attribute"
"go.opentelemetry.io/otel/codes"
"go.opentelemetry.io/otel/internal/trace/parent"
"go.opentelemetry.io/otel/trace"
export "go.opentelemetry.io/otel/sdk/export/trace"
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
"go.opentelemetry.io/otel/sdk/instrumentation"
"go.opentelemetry.io/otel/sdk/internal"
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
"go.opentelemetry.io/otel/sdk/resource"
)
const (
errorTypeKey = attribute.Key("error.type")
errorMessageKey = attribute.Key("error.message")
errorEventName = "error"
)
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
// ReadOnlySpan allows reading information from the data structure underlying a
// trace.Span. It is used in places where reading information from a span is
// necessary but changing the span isn't necessary or allowed.
// TODO: Should we make the methods unexported? The purpose of this interface
// is controlling access to `span` fields, not having multiple implementations.
type ReadOnlySpan interface {
Name() string
SpanContext() trace.SpanContext
Parent() trace.SpanContext
SpanKind() trace.SpanKind
StartTime() time.Time
EndTime() time.Time
Attributes() []attribute.KeyValue
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
Links() []trace.Link
Events() []trace.Event
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
StatusCode() codes.Code
StatusMessage() string
Tracer() trace.Tracer
IsRecording() bool
InstrumentationLibrary() instrumentation.Library
Resource() *resource.Resource
Snapshot() *export.SpanSnapshot
// A private method to prevent users implementing the
// interface and so future additions to it will not
// violate compatibility.
private()
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
}
// ReadWriteSpan exposes the same methods as trace.Span and in addition allows
// reading information from the underlying data structure.
// This interface exposes the union of the methods of trace.Span (which is a
// "write-only" span) and ReadOnlySpan. New methods for writing or reading span
// information should be added under trace.Span or ReadOnlySpan, respectively.
type ReadWriteSpan interface {
trace.Span
ReadOnlySpan
}
// span is an implementation of the OpenTelemetry Span API representing the
// individual component of a trace.
type span struct {
// droppedAttributeCount contains dropped attributes for the events and links.
droppedAttributeCount int64
// mu protects the contents of this span.
mu sync.Mutex
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
// parent holds the parent span of this span as a trace.SpanContext.
parent trace.SpanContext
// spanKind represents the kind of this span as a trace.SpanKind.
spanKind trace.SpanKind
// name is the name of this span.
name string
// startTime is the time at which this span was started.
startTime time.Time
// endTime is the time at which this span was ended. It contains the zero
// value of time.Time until the span is ended.
endTime time.Time
// statusCode represents the status of this span as a codes.Code value.
statusCode codes.Code
// statusMessage represents the status of this span as a string.
statusMessage string
// childSpanCount holds the number of child spans created for this span.
childSpanCount int
// resource contains attributes representing an entity that produced this
// span.
resource *resource.Resource
// instrumentationLibrary defines the instrumentation library used to
// provide instrumentation.
instrumentationLibrary instrumentation.Library
// spanContext holds the SpanContext of this span.
spanContext trace.SpanContext
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
// attributes are capped at configured limit. When the capacity is reached
// an oldest entry is removed to create room for a new entry.
attributes *attributesMap
// messageEvents are stored in FIFO queue capped by configured limit.
messageEvents *evictedQueue
// links are stored in FIFO queue capped by configured limit.
links *evictedQueue
// executionTracerTaskEnd ends the execution tracer span.
executionTracerTaskEnd func()
// tracer is the SDK tracer that created this span.
tracer *tracer
// spanLimits holds the limits to this span.
spanLimits SpanLimits
}
var _ trace.Span = &span{}
// SpanContext returns the SpanContext of this span.
func (s *span) SpanContext() trace.SpanContext {
if s == nil {
return trace.SpanContext{}
}
return s.spanContext
}
// IsRecording returns if this span is being recorded. If this span has ended
// this will return false.
func (s *span) IsRecording() bool {
if s == nil {
return false
}
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
s.mu.Lock()
defer s.mu.Unlock()
return s.endTime.IsZero()
}
// SetStatus sets the status of this span in the form of a code and a
// message. This overrides the existing value of this span's status if one
// exists. Message will be set only if status is error. If this span is not being
// recorded than this method does nothing.
func (s *span) SetStatus(code codes.Code, msg string) {
if !s.IsRecording() {
return
}
s.mu.Lock()
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
s.statusCode = code
if code == codes.Error {
s.statusMessage = msg
}
s.mu.Unlock()
}
// SetAttributes sets attributes of this span.
//
// If a key from attributes already exists the value associated with that key
// will be overwritten with the value contained in attributes.
//
// If this span is not being recorded than this method does nothing.
func (s *span) SetAttributes(attributes ...attribute.KeyValue) {
if !s.IsRecording() {
return
}
s.copyToCappedAttributes(attributes...)
}
// End ends the span. This method does nothing if the span is already ended or
// is not being recorded.
//
// The only SpanOption currently supported is WithTimestamp which will set the
// end time for a Span's life-cycle.
//
// If this method is called while panicking an error event is added to the
// Span before ending it and the panic is continued.
func (s *span) End(options ...trace.SpanOption) {
// Do not start by checking if the span is being recorded which requires
// acquiring a lock. Make a minimal check that the span is not nil.
if s == nil {
return
}
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
// Store the end time as soon as possible to avoid artificially increasing
// the span's duration in case some operation below takes a while.
et := internal.MonotonicEndTime(s.startTime)
// Do relative expensive check now that we have an end time and see if we
// need to do any more processing.
if !s.IsRecording() {
return
}
if recovered := recover(); recovered != nil {
// Record but don't stop the panic.
defer panic(recovered)
s.addEvent(
errorEventName,
trace.WithAttributes(
errorTypeKey.String(typeStr(recovered)),
errorMessageKey.String(fmt.Sprint(recovered)),
),
)
}
if s.executionTracerTaskEnd != nil {
s.executionTracerTaskEnd()
}
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
config := trace.NewSpanConfig(options...)
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
s.mu.Lock()
// Setting endTime to non-zero marks the span as ended and not recording.
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
if config.Timestamp.IsZero() {
s.endTime = et
} else {
s.endTime = config.Timestamp
}
s.mu.Unlock()
sps, ok := s.tracer.provider.spanProcessors.Load().(spanProcessorStates)
mustExportOrProcess := ok && len(sps) > 0
if mustExportOrProcess {
for _, sp := range sps {
sp.sp.OnEnd(s)
}
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
}
}
// RecordError will record err as a span event for this span. An additional call to
// SetStatus is required if the Status of the Span should be set to Error, this method
// does not change the Span status. If this span is not being recorded or err is nil
// than this method does nothing.
func (s *span) RecordError(err error, opts ...trace.EventOption) {
if s == nil || err == nil || !s.IsRecording() {
return
}
opts = append(opts, trace.WithAttributes(
errorTypeKey.String(typeStr(err)),
errorMessageKey.String(err.Error()),
))
s.addEvent(errorEventName, opts...)
}
func typeStr(i interface{}) string {
t := reflect.TypeOf(i)
if t.PkgPath() == "" && t.Name() == "" {
// Likely a builtin type.
return t.String()
}
return fmt.Sprintf("%s.%s", t.PkgPath(), t.Name())
}
// Tracer returns the Tracer that created this span.
func (s *span) Tracer() trace.Tracer {
return s.tracer
}
// AddEvent adds an event with the provided name and options. If this span is
// not being recorded than this method does nothing.
func (s *span) AddEvent(name string, o ...trace.EventOption) {
if !s.IsRecording() {
return
}
s.addEvent(name, o...)
}
func (s *span) addEvent(name string, o ...trace.EventOption) {
c := trace.NewEventConfig(o...)
// Discard over limited attributes
if len(c.Attributes) > s.spanLimits.AttributePerEventCountLimit {
s.addDroppedAttributeCount(len(c.Attributes) - s.spanLimits.AttributePerEventCountLimit)
c.Attributes = c.Attributes[:s.spanLimits.AttributePerEventCountLimit]
}
s.mu.Lock()
defer s.mu.Unlock()
s.messageEvents.add(trace.Event{
Name: name,
Attributes: c.Attributes,
Time: c.Timestamp,
})
}
// SetName sets the name of this span. If this span is not being recorded than
// this method does nothing.
func (s *span) SetName(name string) {
if !s.IsRecording() {
return
}
s.mu.Lock()
defer s.mu.Unlock()
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
s.name = name
}
// Name returns the name of this span.
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
func (s *span) Name() string {
s.mu.Lock()
defer s.mu.Unlock()
return s.name
}
// Name returns the SpanContext of this span's parent span.
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
func (s *span) Parent() trace.SpanContext {
s.mu.Lock()
defer s.mu.Unlock()
return s.parent
}
// SpanKind returns the SpanKind of this span.
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
func (s *span) SpanKind() trace.SpanKind {
s.mu.Lock()
defer s.mu.Unlock()
return s.spanKind
}
// StartTime returns the time this span started.
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
func (s *span) StartTime() time.Time {
s.mu.Lock()
defer s.mu.Unlock()
return s.startTime
}
// EndTime returns the time this span ended. For spans that have not yet
// ended, the returned value will be the zero value of time.Time.
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
func (s *span) EndTime() time.Time {
s.mu.Lock()
defer s.mu.Unlock()
return s.endTime
}
// Attributes returns the attributes of this span.
func (s *span) Attributes() []attribute.KeyValue {
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
s.mu.Lock()
defer s.mu.Unlock()
if s.attributes.evictList.Len() == 0 {
return []attribute.KeyValue{}
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
}
return s.attributes.toKeyValue()
}
// Events returns the links of this span.
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
func (s *span) Links() []trace.Link {
s.mu.Lock()
defer s.mu.Unlock()
if len(s.links.queue) == 0 {
return []trace.Link{}
}
return s.interfaceArrayToLinksArray()
}
// Events returns the events of this span.
func (s *span) Events() []trace.Event {
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
s.mu.Lock()
defer s.mu.Unlock()
if len(s.messageEvents.queue) == 0 {
return []trace.Event{}
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
}
return s.interfaceArrayToMessageEventArray()
}
// StatusCode returns the status code of this span.
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
func (s *span) StatusCode() codes.Code {
s.mu.Lock()
defer s.mu.Unlock()
return s.statusCode
}
// StatusMessage returns the status message of this span.
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
func (s *span) StatusMessage() string {
s.mu.Lock()
defer s.mu.Unlock()
return s.statusMessage
}
// InstrumentationLibrary returns the instrumentation.Library associated with
// the Tracer that created this span.
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
func (s *span) InstrumentationLibrary() instrumentation.Library {
s.mu.Lock()
defer s.mu.Unlock()
return s.instrumentationLibrary
}
// Resource returns the Resource associated with the Tracer that created this
// span.
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
func (s *span) Resource() *resource.Resource {
s.mu.Lock()
defer s.mu.Unlock()
return s.resource
}
func (s *span) addLink(link trace.Link) {
if !s.IsRecording() {
return
}
s.mu.Lock()
defer s.mu.Unlock()
// Discard over limited attributes
if len(link.Attributes) > s.spanLimits.AttributePerLinkCountLimit {
s.addDroppedAttributeCount(len(link.Attributes) - s.spanLimits.AttributePerLinkCountLimit)
link.Attributes = link.Attributes[:s.spanLimits.AttributePerLinkCountLimit]
}
s.links.add(link)
}
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
// Snapshot creates a snapshot representing the current state of the span as an
// export.SpanSnapshot and returns a pointer to it.
func (s *span) Snapshot() *export.SpanSnapshot {
var sd export.SpanSnapshot
s.mu.Lock()
defer s.mu.Unlock()
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
sd.ChildSpanCount = s.childSpanCount
sd.EndTime = s.endTime
sd.HasRemoteParent = s.parent.IsRemote()
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
sd.InstrumentationLibrary = s.instrumentationLibrary
sd.Name = s.name
sd.ParentSpanID = s.parent.SpanID()
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
sd.Resource = s.resource
sd.SpanContext = s.spanContext
sd.SpanKind = s.spanKind
sd.StartTime = s.startTime
sd.StatusCode = s.statusCode
sd.StatusMessage = s.statusMessage
sd.DroppedAttributeCount = int(s.droppedAttributeCount)
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
if s.attributes.evictList.Len() > 0 {
sd.Attributes = s.attributes.toKeyValue()
sd.DroppedAttributeCount += s.attributes.droppedCount
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
}
if len(s.messageEvents.queue) > 0 {
sd.MessageEvents = s.interfaceArrayToMessageEventArray()
sd.DroppedMessageEventCount = s.messageEvents.droppedCount
}
if len(s.links.queue) > 0 {
sd.Links = s.interfaceArrayToLinksArray()
sd.DroppedLinkCount = s.links.droppedCount
}
return &sd
}
func (s *span) interfaceArrayToLinksArray() []trace.Link {
linkArr := make([]trace.Link, 0)
for _, value := range s.links.queue {
linkArr = append(linkArr, value.(trace.Link))
}
return linkArr
}
func (s *span) interfaceArrayToMessageEventArray() []trace.Event {
messageEventArr := make([]trace.Event, 0)
for _, value := range s.messageEvents.queue {
messageEventArr = append(messageEventArr, value.(trace.Event))
}
return messageEventArr
}
func (s *span) copyToCappedAttributes(attributes ...attribute.KeyValue) {
s.mu.Lock()
defer s.mu.Unlock()
for _, a := range attributes {
// Ensure attributes conform to the specification:
// https://github.com/open-telemetry/opentelemetry-specification/blob/v1.0.1/specification/common/common.md#attributes
if a.Valid() {
s.attributes.add(a)
}
}
}
func (s *span) addChild() {
if !s.IsRecording() {
return
}
s.mu.Lock()
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
s.childSpanCount++
s.mu.Unlock()
}
func (s *span) addDroppedAttributeCount(delta int) {
atomic.AddInt64(&s.droppedAttributeCount, int64(delta))
}
func (*span) private() {}
func startSpanInternal(ctx context.Context, tr *tracer, name string, o *trace.SpanConfig) *span {
span := &span{}
provider := tr.provider
// If told explicitly to make this a new root use a zero value SpanContext
// as a parent which contains an invalid trace ID and is not remote.
var psc trace.SpanContext
if !o.NewRoot {
psc = parent.SpanContext(ctx)
}
// If there is a valid parent trace ID, use it to ensure the continuity of
// the trace. Always generate a new span ID so other components can rely
// on a unique span ID, even if the Span is non-recording.
var tid trace.TraceID
var sid trace.SpanID
if !psc.TraceID().IsValid() {
tid, sid = provider.idGenerator.NewIDs(ctx)
} else {
tid = psc.TraceID()
sid = provider.idGenerator.NewSpanID(ctx, tid)
}
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
spanLimits := provider.spanLimits
span.attributes = newAttributesMap(spanLimits.AttributeCountLimit)
span.messageEvents = newEvictedQueue(spanLimits.EventCountLimit)
span.links = newEvictedQueue(spanLimits.LinkCountLimit)
span.spanLimits = spanLimits
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
samplingResult := provider.sampler.ShouldSample(SamplingParameters{
ParentContext: psc,
TraceID: tid,
Name: name,
HasRemoteParent: psc.IsRemote(),
Kind: o.SpanKind,
Attributes: o.Attributes,
Links: o.Links,
})
scc := trace.SpanContextConfig{
TraceID: tid,
SpanID: sid,
TraceState: samplingResult.Tracestate,
}
if isSampled(samplingResult) {
scc.TraceFlags = psc.TraceFlags() | trace.FlagsSampled
} else {
scc.TraceFlags = psc.TraceFlags() &^ trace.FlagsSampled
}
span.spanContext = trace.NewSpanContext(scc)
if !isRecording(samplingResult) {
return span
}
startTime := o.Timestamp
if startTime.IsZero() {
startTime = time.Now()
}
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
span.startTime = startTime
span.spanKind = trace.ValidateSpanKind(o.SpanKind)
span.name = name
span.parent = psc
span.resource = provider.resource
Add RO/RW span interfaces (#1360) * Store span data directly in the span - Nesting only some of a span's data in a `data` field (with the rest of the data living direclty in the `span` struct) is confusing. - export.SpanData is meant to be an immutable *snapshot* of a span, not the "authoritative" state of the span. - Refactor attributesMap.toSpanData into toKeyValue and make it return a []label.KeyValue which is clearer than modifying a struct passed to the function. - Read droppedCount from the attributesMap as a separate operation instead of setting it from within attributesMap.toSpanData. - Set a span's end time in the span itself rather than in the SpanData to allow reading the span's end time after a span has ended. - Set a span's end time as soon as possible within span.End so that we don't influence the span's end time with operations such as fetching span processors and generating span data. - Remove error handling for uninitialized spans. This check seems to be necessary only because we used to have an *export.SpanData field which could be nil. Now that we no longer have this field I think we can safely remove the check. The error isn't used anywhere else so remove it, too. * Store parent as trace.SpanContext The spec requires that the parent field of a Span be a Span, a SpanContext or null. Rather than extracting the parent's span ID from the trace.SpanContext which we get from the tracer, store the trace.SpanContext as is and explicitly extract the parent's span ID where necessary. * Add ReadOnlySpan interface Use this interface instead of export.SpanData in places where reading information from a span is necessary. Use export.SpanData only when exporting spans. * Add ReadWriteSpan interface Use this interface instead of export.SpanData in places where it is necessary to read information from a span and write to it at the same time. * Rename export.SpanData to SpanSnapshot SpanSnapshot represents the nature of this type as well as its intended use more accurately. Clarify the purpose of SpanSnapshot in the docs and emphasize what should and should not be done with it. * Rephrase attributesMap doc comment "refreshes" is wrong for plural ("updates"). * Refactor span.End() - Improve accuracy of span duration. Record span end time ASAP. We want to measure a user operation as accurately as possible, which means we want to mark the end time of a span as soon as possible after span.End() is called. Any operations we do inside span.End() before storing the end time affect the total duration of the span, and although these operations are rather fast at the moment they still seem to affect the duration of the span by "artificially" adding time between the start and end timestamps. This is relevant only in cases where the end time isn't explicitly specified. - Remove redundant idempotence check. Now that IsRecording() is based on the value of span.endTime, IsRecording() will always return false after span.End() had been called because span.endTime won't be zero. This means we no longer need span.endOnce. - Improve TestEndSpanTwice so that it also ensures subsequent calls to span.End() don't modify the span's end time. * Update changelog Co-authored-by: Tyler Yahn <codingalias@gmail.com> Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2020-12-11 07:15:44 +02:00
span.instrumentationLibrary = tr.instrumentationLibrary
span.SetAttributes(samplingResult.Attributes...)
return span
}
func isRecording(s SamplingResult) bool {
return s.Decision == RecordOnly || s.Decision == RecordAndSample
}
func isSampled(s SamplingResult) bool {
return s.Decision == RecordAndSample
}