2020-03-25 14:47:17 -07:00
|
|
|
// Copyright The OpenTelemetry Authors
|
2020-03-06 22:01:02 -08:00
|
|
|
//
|
|
|
|
// 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.
|
|
|
|
|
2020-11-04 18:10:58 +01:00
|
|
|
package otlp // import "go.opentelemetry.io/otel/exporters/otlp"
|
2020-03-06 22:01:02 -08:00
|
|
|
|
2020-10-11 19:57:50 -07:00
|
|
|
// This code was based on
|
|
|
|
// contrib.go.opencensus.io/exporter/ocagent/connection.go
|
|
|
|
|
2020-03-06 22:01:02 -08:00
|
|
|
import (
|
|
|
|
"context"
|
|
|
|
"errors"
|
|
|
|
"sync"
|
|
|
|
|
|
|
|
"google.golang.org/grpc"
|
|
|
|
|
2020-08-10 19:55:52 -07:00
|
|
|
colmetricpb "go.opentelemetry.io/otel/exporters/otlp/internal/opentelemetry-proto-gen/collector/metrics/v1"
|
|
|
|
coltracepb "go.opentelemetry.io/otel/exporters/otlp/internal/opentelemetry-proto-gen/collector/trace/v1"
|
2020-03-13 11:42:20 -07:00
|
|
|
"go.opentelemetry.io/otel/exporters/otlp/internal/transform"
|
2020-11-12 16:28:32 +01:00
|
|
|
"go.opentelemetry.io/otel/metric"
|
2020-03-13 11:42:20 -07:00
|
|
|
metricsdk "go.opentelemetry.io/otel/sdk/export/metric"
|
2020-06-22 22:59:51 -07:00
|
|
|
"go.opentelemetry.io/otel/sdk/export/metric/aggregation"
|
2020-03-13 11:42:20 -07:00
|
|
|
tracesdk "go.opentelemetry.io/otel/sdk/export/trace"
|
2020-03-06 22:01:02 -08:00
|
|
|
)
|
|
|
|
|
2020-10-11 19:57:50 -07:00
|
|
|
// Exporter is an OpenTelemetry exporter. It exports both traces and metrics
|
|
|
|
// from OpenTelemetry instrumented to code using OpenTelemetry protocol
|
|
|
|
// buffers to a configurable receiver.
|
2020-03-06 22:01:02 -08:00
|
|
|
type Exporter struct {
|
|
|
|
// mu protects the non-atomic and non-channel variables
|
|
|
|
mu sync.RWMutex
|
2020-03-13 11:42:20 -07:00
|
|
|
// senderMu protects the concurrent unsafe sends on the shared gRPC client connection.
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
senderMu sync.Mutex
|
|
|
|
started bool
|
|
|
|
traceExporter coltracepb.TraceServiceClient
|
|
|
|
metricExporter colmetricpb.MetricsServiceClient
|
|
|
|
cc *grpcConnection
|
|
|
|
|
|
|
|
startOnce sync.Once
|
|
|
|
stopOnce sync.Once
|
|
|
|
|
|
|
|
exportKindSelector metricsdk.ExportKindSelector
|
2020-03-06 22:01:02 -08:00
|
|
|
}
|
|
|
|
|
2020-09-09 10:19:03 -07:00
|
|
|
var _ tracesdk.SpanExporter = (*Exporter)(nil)
|
2020-03-13 11:42:20 -07:00
|
|
|
var _ metricsdk.Exporter = (*Exporter)(nil)
|
2020-03-06 22:01:02 -08:00
|
|
|
|
2020-09-11 17:14:25 -04:00
|
|
|
// newConfig initializes a config struct with default values and applies
|
|
|
|
// any ExporterOptions provided.
|
|
|
|
func newConfig(opts ...ExporterOption) config {
|
|
|
|
cfg := config{
|
|
|
|
grpcServiceConfig: DefaultGRPCServiceConfig,
|
2020-11-10 07:44:42 -08:00
|
|
|
|
|
|
|
// Note: the default ExportKindSelector is specified
|
|
|
|
// as Cumulative:
|
|
|
|
// https://github.com/open-telemetry/opentelemetry-specification/issues/731
|
|
|
|
exportKindSelector: metricsdk.CumulativeExportKindSelector(),
|
2020-09-11 17:14:25 -04:00
|
|
|
}
|
2020-03-06 22:01:02 -08:00
|
|
|
for _, opt := range opts {
|
2020-09-11 17:14:25 -04:00
|
|
|
opt(&cfg)
|
2020-03-06 22:01:02 -08:00
|
|
|
}
|
2020-09-11 17:14:25 -04:00
|
|
|
return cfg
|
2020-03-06 22:01:02 -08:00
|
|
|
}
|
|
|
|
|
2020-09-11 17:14:25 -04:00
|
|
|
// NewExporter constructs a new Exporter and starts it.
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
func NewExporter(ctx context.Context, opts ...ExporterOption) (*Exporter, error) {
|
2020-03-06 22:01:02 -08:00
|
|
|
exp := NewUnstartedExporter(opts...)
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
if err := exp.Start(ctx); err != nil {
|
2020-03-06 22:01:02 -08:00
|
|
|
return nil, err
|
|
|
|
}
|
|
|
|
return exp, nil
|
|
|
|
}
|
|
|
|
|
2020-09-11 17:14:25 -04:00
|
|
|
// NewUnstartedExporter constructs a new Exporter and does not start it.
|
2020-03-06 22:01:02 -08:00
|
|
|
func NewUnstartedExporter(opts ...ExporterOption) *Exporter {
|
|
|
|
e := new(Exporter)
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
cfg := newConfig(opts...)
|
|
|
|
e.exportKindSelector = cfg.exportKindSelector
|
|
|
|
e.cc = newGRPCConnection(cfg, e.handleNewConnection)
|
2020-03-06 22:01:02 -08:00
|
|
|
return e
|
|
|
|
}
|
|
|
|
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
func (e *Exporter) handleNewConnection(cc *grpc.ClientConn) error {
|
|
|
|
e.mu.Lock()
|
|
|
|
defer e.mu.Unlock()
|
|
|
|
if cc != nil {
|
|
|
|
e.metricExporter = colmetricpb.NewMetricsServiceClient(cc)
|
|
|
|
e.traceExporter = coltracepb.NewTraceServiceClient(cc)
|
|
|
|
} else {
|
|
|
|
e.metricExporter = nil
|
|
|
|
e.traceExporter = nil
|
|
|
|
}
|
|
|
|
return nil
|
|
|
|
}
|
|
|
|
|
2020-03-06 22:01:02 -08:00
|
|
|
var (
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
errNoClient = errors.New("no client")
|
|
|
|
errAlreadyStarted = errors.New("already started")
|
|
|
|
errDisconnected = errors.New("exporter disconnected")
|
2020-03-06 22:01:02 -08:00
|
|
|
)
|
|
|
|
|
|
|
|
// Start dials to the collector, establishing a connection to it. It also
|
|
|
|
// initiates the Config and Trace services by sending over the initial
|
|
|
|
// messages that consist of the node identifier. Start invokes a background
|
|
|
|
// connector that will reattempt connections to the collector periodically
|
|
|
|
// if the connection dies.
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
func (e *Exporter) Start(ctx context.Context) error {
|
2020-03-06 22:01:02 -08:00
|
|
|
var err = errAlreadyStarted
|
|
|
|
e.startOnce.Do(func() {
|
|
|
|
e.mu.Lock()
|
|
|
|
e.started = true
|
|
|
|
e.mu.Unlock()
|
|
|
|
|
|
|
|
err = nil
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
e.cc.startConnection(ctx)
|
2020-03-06 22:01:02 -08:00
|
|
|
})
|
|
|
|
|
|
|
|
return err
|
|
|
|
}
|
|
|
|
|
2020-09-09 10:19:03 -07:00
|
|
|
// Shutdown closes all connections and releases resources currently being used
|
|
|
|
// by the exporter. If the exporter is not started this does nothing.
|
|
|
|
func (e *Exporter) Shutdown(ctx context.Context) error {
|
2020-03-06 22:01:02 -08:00
|
|
|
e.mu.RLock()
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
cc := e.cc
|
2020-03-06 22:01:02 -08:00
|
|
|
started := e.started
|
|
|
|
e.mu.RUnlock()
|
|
|
|
|
|
|
|
if !started {
|
|
|
|
return nil
|
|
|
|
}
|
|
|
|
|
|
|
|
var err error
|
|
|
|
|
Some cleanups in otlp exporter (#1350)
* Drop WorkerCount option
This is not a good option - the user isn't likely to know how many
worker goroutines is optimal. This should be something that an
exporter should figure out itself. The second problem with the option
is that it is specific to the metric transformation from SDK export
format into protobuf. When the exporter starts supporting other
protocols (HTTP/JSON for example), this option may be of no use. So
the option should rather belong to the protocol, not to the
exporter. Currently both mean the same, but later they will be
separated, and this option breaks the separation.
* Make stop channel a typical signalling channel
Signalling channels are idiomatically defined as chan struct{}, so
let's make it so, to avoid confusion about the meaning of the bool
type.
* Close a race when grpc connection is closed multiple times
If several goroutines call Shutdown at the same time, then the
following scenario is possible:
goroutine A locks a mutex, reads a started member, unlocks the mutex
and gets preempted
goroutine B locks a mutex, reads a started member, unlocks the mutex
and gets preempted
goroutine A does not return early in the "if !started" conditional and
continues to close the connection and execute the rest of the function
(where it finally sets the started member to false), gets preempted
goroutine B also does not return early, because it got a copy of
started before goroutine A set it to false, so it tries to close the
connection again.
* Update changelog
2020-11-20 03:03:25 +01:00
|
|
|
e.stopOnce.Do(func() {
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
// Clean things up before checking this error.
|
|
|
|
err = cc.shutdown(ctx)
|
2020-03-06 22:01:02 -08:00
|
|
|
|
Some cleanups in otlp exporter (#1350)
* Drop WorkerCount option
This is not a good option - the user isn't likely to know how many
worker goroutines is optimal. This should be something that an
exporter should figure out itself. The second problem with the option
is that it is specific to the metric transformation from SDK export
format into protobuf. When the exporter starts supporting other
protocols (HTTP/JSON for example), this option may be of no use. So
the option should rather belong to the protocol, not to the
exporter. Currently both mean the same, but later they will be
separated, and this option breaks the separation.
* Make stop channel a typical signalling channel
Signalling channels are idiomatically defined as chan struct{}, so
let's make it so, to avoid confusion about the meaning of the bool
type.
* Close a race when grpc connection is closed multiple times
If several goroutines call Shutdown at the same time, then the
following scenario is possible:
goroutine A locks a mutex, reads a started member, unlocks the mutex
and gets preempted
goroutine B locks a mutex, reads a started member, unlocks the mutex
and gets preempted
goroutine A does not return early in the "if !started" conditional and
continues to close the connection and execute the rest of the function
(where it finally sets the started member to false), gets preempted
goroutine B also does not return early, because it got a copy of
started before goroutine A set it to false, so it tries to close the
connection again.
* Update changelog
2020-11-20 03:03:25 +01:00
|
|
|
// At this point we can change the state variable started
|
|
|
|
e.mu.Lock()
|
|
|
|
e.started = false
|
|
|
|
e.mu.Unlock()
|
|
|
|
})
|
2020-03-06 22:01:02 -08:00
|
|
|
|
|
|
|
return err
|
|
|
|
}
|
|
|
|
|
2020-03-13 11:42:20 -07:00
|
|
|
// Export implements the "go.opentelemetry.io/otel/sdk/export/metric".Exporter
|
2020-04-15 12:04:44 -07:00
|
|
|
// interface. It transforms and batches metric Records into OTLP Metrics and
|
|
|
|
// transmits them to the configured collector.
|
2020-05-18 17:44:28 -07:00
|
|
|
func (e *Exporter) Export(parent context.Context, cps metricsdk.CheckpointSet) error {
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
ctx, cancel := e.cc.contextWithStop(parent)
|
2020-04-15 12:04:44 -07:00
|
|
|
defer cancel()
|
2020-03-13 11:42:20 -07:00
|
|
|
|
Some cleanups in otlp exporter (#1350)
* Drop WorkerCount option
This is not a good option - the user isn't likely to know how many
worker goroutines is optimal. This should be something that an
exporter should figure out itself. The second problem with the option
is that it is specific to the metric transformation from SDK export
format into protobuf. When the exporter starts supporting other
protocols (HTTP/JSON for example), this option may be of no use. So
the option should rather belong to the protocol, not to the
exporter. Currently both mean the same, but later they will be
separated, and this option breaks the separation.
* Make stop channel a typical signalling channel
Signalling channels are idiomatically defined as chan struct{}, so
let's make it so, to avoid confusion about the meaning of the bool
type.
* Close a race when grpc connection is closed multiple times
If several goroutines call Shutdown at the same time, then the
following scenario is possible:
goroutine A locks a mutex, reads a started member, unlocks the mutex
and gets preempted
goroutine B locks a mutex, reads a started member, unlocks the mutex
and gets preempted
goroutine A does not return early in the "if !started" conditional and
continues to close the connection and execute the rest of the function
(where it finally sets the started member to false), gets preempted
goroutine B also does not return early, because it got a copy of
started before goroutine A set it to false, so it tries to close the
connection again.
* Update changelog
2020-11-20 03:03:25 +01:00
|
|
|
// Hardcode the number of worker goroutines to 1. We later will
|
|
|
|
// need to see if there's a way to adjust that number for longer
|
|
|
|
// running operations.
|
|
|
|
rms, err := transform.CheckpointSet(ctx, e, cps, 1)
|
2020-04-15 12:04:44 -07:00
|
|
|
if err != nil {
|
|
|
|
return err
|
2020-03-13 11:42:20 -07:00
|
|
|
}
|
|
|
|
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
if !e.cc.connected() {
|
2020-04-15 12:04:44 -07:00
|
|
|
return errDisconnected
|
2020-03-13 11:42:20 -07:00
|
|
|
}
|
|
|
|
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
err = func() error {
|
2020-03-13 11:42:20 -07:00
|
|
|
e.senderMu.Lock()
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
defer e.senderMu.Unlock()
|
|
|
|
if e.metricExporter == nil {
|
|
|
|
return errNoClient
|
|
|
|
}
|
|
|
|
_, err := e.metricExporter.Export(e.cc.contextWithMetadata(ctx), &colmetricpb.ExportMetricsServiceRequest{
|
2020-04-15 12:04:44 -07:00
|
|
|
ResourceMetrics: rms,
|
2020-03-13 11:42:20 -07:00
|
|
|
})
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
return err
|
|
|
|
}()
|
|
|
|
if err != nil {
|
|
|
|
e.cc.setStateDisconnected(err)
|
2020-03-13 11:42:20 -07:00
|
|
|
}
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
return err
|
2020-03-13 11:42:20 -07:00
|
|
|
}
|
|
|
|
|
2020-10-11 19:57:50 -07:00
|
|
|
// ExportKindFor reports back to the OpenTelemetry SDK sending this Exporter
|
2020-11-10 07:44:42 -08:00
|
|
|
// metric telemetry that it needs to be provided in a cumulative format.
|
2020-11-12 16:28:32 +01:00
|
|
|
func (e *Exporter) ExportKindFor(desc *metric.Descriptor, kind aggregation.Kind) metricsdk.ExportKind {
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
return e.exportKindSelector.ExportKindFor(desc, kind)
|
2020-06-22 22:59:51 -07:00
|
|
|
}
|
|
|
|
|
2020-12-11 06:15:44 +01:00
|
|
|
// ExportSpans exports a batch of SpanSnapshot.
|
|
|
|
func (e *Exporter) ExportSpans(ctx context.Context, ss []*tracesdk.SpanSnapshot) error {
|
|
|
|
return e.uploadTraces(ctx, ss)
|
2020-03-13 11:42:20 -07:00
|
|
|
}
|
|
|
|
|
2020-12-11 06:15:44 +01:00
|
|
|
func (e *Exporter) uploadTraces(ctx context.Context, ss []*tracesdk.SpanSnapshot) error {
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
ctx, cancel := e.cc.contextWithStop(ctx)
|
|
|
|
defer cancel()
|
|
|
|
|
|
|
|
if !e.cc.connected() {
|
2020-09-09 10:19:03 -07:00
|
|
|
return nil
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
}
|
2020-03-06 22:01:02 -08:00
|
|
|
|
2020-12-11 06:15:44 +01:00
|
|
|
protoSpans := transform.SpanData(ss)
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
if len(protoSpans) == 0 {
|
|
|
|
return nil
|
|
|
|
}
|
2020-03-06 22:01:02 -08:00
|
|
|
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
err := func() error {
|
2020-03-06 22:01:02 -08:00
|
|
|
e.senderMu.Lock()
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
defer e.senderMu.Unlock()
|
|
|
|
if e.traceExporter == nil {
|
|
|
|
return errNoClient
|
|
|
|
}
|
|
|
|
_, err := e.traceExporter.Export(e.cc.contextWithMetadata(ctx), &coltracepb.ExportTraceServiceRequest{
|
2020-03-06 22:01:02 -08:00
|
|
|
ResourceSpans: protoSpans,
|
|
|
|
})
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
return err
|
|
|
|
}()
|
|
|
|
if err != nil {
|
|
|
|
e.cc.setStateDisconnected(err)
|
2020-03-06 22:01:02 -08:00
|
|
|
}
|
Another batch of cleanups in otlp exporter (#1357)
* Move connection logic into grpcConnection object
If we will need to maintain more than one connection in future, this
splitting off will come in handy.
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
* Make another channel a signal channel
There is another channel that serves as a one-time signal, where
channel's data type does not matter.
* Reorder and document connection members
This is to make clear that the lock is guarding only the connection
since it can be changed by multiple goroutines, and other members are
either atomic or read-only.
* Move stop signal into connection
The stop channel was rather useless on the exporter side - the primary
reason for existence of this channel is to stop a background
reconnecting goroutine. Since the goroutine lives entirely within
grpcConnection object, move the stop channel here. Also expose a
function to unify the stop channel with the context cancellation, so
exporter can use it without knowing anything about stop channels.
Also make export functions a bit more consistent.
* Do not run reconnection routine when being stopped too
It's possible that both disconnected channel and stop channel will be
triggered around the same time, so the goroutine is as likely to start
reconnecting as to return from the goroutine. Make sure we return if
the stop channel is closed.
* Nil clients on connection error
Set clients to nil on connection error, so we don't try to send the
data over a bad connection, but return a "no client" error
immediately.
* Do not call new connection handler within critical section
It's rather risky to call a callback coming from outside within a
critical section. Move it out.
* Add context parameter to connection routines
Connecting to the collector may also take its time, so it can be
useful in some cases to pass a context with a deadline. Currently we
just pass a background context, so this commit does not really change
any behavior. The follow-up commits will make a use of it, though.
* Add context parameter to NewExporter and Start
It makes it possible to limit the time spent on connecting to the
collector.
* Stop connecting on shutdown
Dialling to grpc service ignored the closing of the stop channel, but
this can be easily changed.
* Close connection after background is shut down
That way we can make sure that there won't be a window between closing
a connection and waiting for the background goroutine to return, where
the new connection could be established.
* Remove unnecessary nil check
This member is never nil, unless the Exporter is created like
&Exporter{}, which is not a thing we support anyway.
* Update changelog
Co-authored-by: Stefan Prisca <stefan.prisca@gmail.com>
2020-11-24 20:50:05 +01:00
|
|
|
return err
|
2020-03-06 22:01:02 -08:00
|
|
|
}
|