1
0
mirror of https://github.com/open-telemetry/opentelemetry-go.git synced 2025-12-01 23:12:29 +02:00
Commit Graph

15 Commits

Author SHA1 Message Date
Nathan Baulch
506a9baf5e Fix typos (#5763) 2024-09-09 08:53:15 +02:00
Tyler Yahn
002c0a4c03 Move log.Processor.Enabled to independent FilterProcessor interfaced type (#5692)
Closes #5425 

Our current log `Processor` interface contains more functionality than
the [OTel
spec](https://github.com/open-telemetry/opentelemetry-specification/blob/main/specification/logs/sdk.md#logrecordprocessor-operations).
The additional functionality allows processors to report back to the API
if a Record should be constructed and emitted or not, which is quite
helpful[^1][^2][^3][^4][^5].

This removes the `Enabled` method from the `Processor` type. It adds
this functionality a new optional and experimental `FilterProcessor`
interface type. The logger and provider are updated to check for this
optional interface to be implemented with the configured processors and
uses them to back the `Logger.Enabled` method, preserving existing
functionality.

By making this change:

- The `Processor` interface is now compliant with the OTel spec and does
not contain any additional unspecified behavior.
- All `Processor` implementations are no longer required to implement an
`Enabled` method. The default, when they do not implement this method,
is to assume they are enabled.

### Benchmark

```terminal
goos: linux
goarch: amd64
pkg: go.opentelemetry.io/otel/sdk/log
cpu: Intel(R) Core(TM) i7-8550U CPU @ 1.80GHz
                │   old.txt    │              new7.txt               │
                │    sec/op    │   sec/op     vs base                │
LoggerEnabled-8   133.30n ± 3%   32.36n ± 3%  -75.72% (p=0.000 n=10)

                │  old.txt   │            new7.txt            │
                │    B/op    │    B/op     vs base            │
LoggerEnabled-8   0.000 ± 0%   0.000 ± 0%  ~ (p=1.000 n=10) ¹
¹ all samples are equal

                │  old.txt   │            new7.txt            │
                │ allocs/op  │ allocs/op   vs base            │
LoggerEnabled-8   0.000 ± 0%   0.000 ± 0%  ~ (p=1.000 n=10) ¹
¹ all samples are equal
```

This is a significant performance improvement due to the `Record` no
longer being converted from the API version to the SDK version.

[^1]: https://pkg.go.dev/go.opentelemetry.io/contrib/processors/minsev
[^2]:
https://pkg.go.dev/go.opentelemetry.io/otel/sdk/log#BatchProcessor.Enabled
[^3]:
https://pkg.go.dev/go.opentelemetry.io/otel/sdk/log#SimpleProcessor.Enabled
[^4]:
af75717ac4/bridges/otelslog/handler.go (L206-L211)
[^5]:
d0309ddd8c/bridges/otelzap/core.go (L142-L146)

---------

Co-authored-by: Robert Pająk <pellared@hotmail.com>
Co-authored-by: Sam Xie <sam@samxie.me>
2024-08-22 09:12:23 -07:00
Tyler Yahn
c4ebcaa08d Ensure exported struct in sdk/log are not comparable (#5693)
Ensure the `Record` and `Provider` continue to be non-comparable.

Restrict `BatchProcessor` and `SimpleProcessor` to be non-comparable.

There is no obvious reason an end-user will need to compare these types,
and we want to keep the possibility of changing the internals without
changing this behavior.

if comparability is required by end-users in the future we can add
`Equal(other T)` methods in the future.
2024-08-09 07:31:42 -07:00
Robert Pająk
69e235829e sdk/log: SimpleProcessor synchronizes OnEmit calls (#5666)
From
https://github.com/open-telemetry/opentelemetry-specification/blob/main/specification/logs/sdk.md#export:

> `Export` will never be called concurrently for the same exporter
instance.

From our SDK perspective this change will make
https://pkg.go.dev/go.opentelemetry.io/otel/exporters/stdout/stdoutlog
concurrent-safe when used with the simple processor. Before the change,
there can be multiple goroutines calling `Write` in parallel (via
`json.Encoder.Encode`).

### Remarks

[Maybe we should simply state that "whole" exporter implementation does
not need to be
concurrent-safe?](https://github.com/open-telemetry/opentelemetry-specification/pull/4173#discussion_r1699958752)
However:
1. we would need to make complex changes in `BatchExporter` to
synchronize the `Export`, `Shutdown`, `ForceFlush` calls
2. we would need to update all exporters (remove synchronization) and
simple exporter (add locks to `Shutdown`, `ForceFlush`)
3. I am 100% not sure if this would be compliant with the specification
- I think it would be complaint because we would simply give stronger
safety-measures


We should probably discuss it separately, but I wanted to highlight my
though process.
Even if we decide that simple and batch processors to synchronize all
calls then I would prefer to address it in a separate PR.

Related spec clarification PR:
-
https://github.com/open-telemetry/opentelemetry-specification/pull/4173
2024-08-09 08:40:39 +02:00
Robert Pająk
f079b03364 sdk/log: SimpleProcessor to not panic for zero value (#5665) 2024-08-04 12:35:58 +02:00
Robert Pająk
8e8ad092cc sdk/log: Processor.OnEmit accetps a Record pointer (#5636)
## What

Change `Processor.OnEmit` methods to accept a record pointer.

## Why

Fixes https://github.com/open-telemetry/opentelemetry-go/issues/5219

This would be specification compliant according to discussions around
https://github.com/open-telemetry/opentelemetry-specification/pull/4067

This is inline of how processors Go span processors works and how log
processors work in other languages.

If the performance (an additional heap allocation during log processing)
would occur to be a significant problem for some users, we have at few
possible solutions:

1. Utilize PGO which may also lead to decreasing heap allocations
(sources:
https://landontclipp.github.io/blog/2023/08/25/profile-guided-optimizations-in-go/#devirtualization,
https://andrewwphillips.github.io/blog/pgo.html). Currently it does not
but I expect it may change in future.
2. Improve the Go compilers escape analysis (related to previous point)
3. introduce new "chaining processor" which can be also desirable in
other languages

## Benchstat

`old` is from `main`.
`new` is from current branch.

`new-pgo` is from current branch with PGO optimization. I first run
benchmarks to generate a CPU profile using `go test -run=^$ -bench=.
-count=10 -cpuprofile default.pgo` and then I rerun the tests with PGO.
Currently, the number of heap allocations is the same.

```
goos: linux
goarch: amd64
pkg: go.opentelemetry.io/otel/sdk/log
cpu: Intel(R) Core(TM) i9-10885H CPU @ 2.40GHz
                                  │    old.txt    │                new.txt                │              new-pgo.txt              │
                                  │    sec/op     │    sec/op     vs base                 │    sec/op     vs base                 │
BatchProcessorOnEmit-16              402.7n ± 18%   382.0n ±  7%         ~ (p=0.247 n=10)   376.7n ± 14%         ~ (p=0.210 n=10)
Processor/Simple-16                  350.9n ±  9%   782.5n ±  6%  +123.00% (p=0.000 n=10)   755.1n ±  5%  +115.19% (p=0.000 n=10)
Processor/Batch-16                   1.333µ ± 15%   1.497µ ± 11%   +12.27% (p=0.000 n=10)   1.528µ ±  8%   +14.63% (p=0.000 n=10)
Processor/SetTimestampSimple-16      329.5n ± 15%   711.6n ±  4%  +115.93% (p=0.000 n=10)   721.9n ±  5%  +119.04% (p=0.000 n=10)
Processor/SetTimestampBatch-16       1.163µ ±  2%   1.524µ ±  3%   +31.03% (p=0.000 n=10)   1.461µ ±  5%   +25.57% (p=0.000 n=10)
Processor/AddAttributesSimple-16     408.7n ±  3%   810.1n ±  4%   +98.21% (p=0.000 n=10)   830.1n ±  4%  +103.11% (p=0.000 n=10)
Processor/AddAttributesBatch-16      1.270µ ±  2%   1.623µ ±  4%   +27.71% (p=0.000 n=10)   1.597µ ±  7%   +25.66% (p=0.000 n=10)
Processor/SetAttributesSimple-16     436.2n ± 10%   796.1n ±  3%   +82.50% (p=0.000 n=10)   817.6n ±  4%   +87.43% (p=0.000 n=10)
Processor/SetAttributesBatch-16      1.202µ ±  2%   1.552µ ±  2%   +29.06% (p=0.000 n=10)   1.659µ ± 11%   +37.96% (p=0.000 n=10)
LoggerNewRecord/5_attributes-16      366.6n ±  3%   363.7n ±  7%         ~ (p=0.952 n=10)   426.2n ±  7%   +16.27% (p=0.000 n=10)
LoggerNewRecord/10_attributes-16     1.711µ ±  2%   1.909µ ± 18%   +11.57% (p=0.000 n=10)   2.077µ ± 10%   +21.39% (p=0.000 n=10)
LoggerProviderLogger-16              650.1n ±  4%   690.1n ±  8%    +6.15% (p=0.019 n=10)   737.6n ± 13%   +13.47% (p=0.004 n=10)
WalkAttributes/1_attributes-16       5.264n ± 12%   5.510n ±  8%         ~ (p=0.812 n=10)   5.865n ±  5%   +11.41% (p=0.011 n=10)
WalkAttributes/10_attributes-16      5.440n ±  8%   5.881n ±  7%    +8.12% (p=0.004 n=10)   6.104n ±  7%   +12.21% (p=0.005 n=10)
WalkAttributes/100_attributes-16     5.403n ±  9%   5.894n ±  9%    +9.10% (p=0.029 n=10)   5.783n ±  6%         ~ (p=0.052 n=10)
WalkAttributes/1000_attributes-16    5.196n ±  4%   5.860n ±  8%   +12.79% (p=0.000 n=10)   5.981n ± 13%   +15.13% (p=0.002 n=10)
SetAddAttributes/SetAttributes-16    181.2n ± 14%   208.1n ± 12%   +14.85% (p=0.005 n=10)   209.9n ± 11%   +15.87% (p=0.007 n=10)
SetAddAttributes/AddAttributes-16    156.7n ± 14%   161.1n ± 16%         ~ (p=0.190 n=10)   165.5n ± 15%         ~ (p=0.315 n=10)
SimpleProcessorOnEmit-16            11.775n ± 10%   9.027n ± 17%   -23.33% (p=0.000 n=10)   9.389n ± 18%   -20.26% (p=0.002 n=10)
geomean                              169.1n         209.6n         +23.98%                  215.5n         +27.48%

                        │     old.txt     │               new.txt                │              new-pgo.txt              │
                        │       B/s       │     B/s       vs base                │      B/s       vs base                │
BatchProcessorOnEmit-16   1004.39Mi ± 15%   79.88Mi ± 7%  -92.05% (p=0.000 n=10)   81.06Mi ± 12%  -91.93% (p=0.000 n=10)

                                  │   old.txt    │                new.txt                 │               new-pgo.txt               │
                                  │     B/op     │    B/op      vs base                   │     B/op      vs base                   │
BatchProcessorOnEmit-16             0.000 ± 0%      0.000 ± 0%         ~ (p=1.000 n=10) ¹    0.000 ±  0%         ~ (p=1.000 n=10) ¹
Processor/Simple-16                   0.0 ± 0%      417.0 ± 0%         ? (p=0.000 n=10)      417.0 ±  0%         ? (p=0.000 n=10)
Processor/Batch-16                  621.5 ± 2%     1057.5 ± 1%   +70.15% (p=0.000 n=10)     1064.5 ±  1%   +71.28% (p=0.000 n=10)
Processor/SetTimestampSimple-16       0.0 ± 0%      417.0 ± 0%         ? (p=0.000 n=10)      418.0 ±  0%         ? (p=0.000 n=10)
Processor/SetTimestampBatch-16      626.5 ± 3%     1049.5 ± 1%   +67.52% (p=0.000 n=10)     1057.5 ±  2%   +68.79% (p=0.000 n=10)
Processor/AddAttributesSimple-16      0.0 ± 0%      417.0 ± 0%         ? (p=0.000 n=10)      417.0 ±  0%         ? (p=0.000 n=10)
Processor/AddAttributesBatch-16     616.5 ± 3%     1053.0 ± 2%   +70.80% (p=0.000 n=10)     1048.5 ±  2%   +70.07% (p=0.000 n=10)
Processor/SetAttributesSimple-16    48.00 ± 0%     466.00 ± 0%  +870.83% (p=0.000 n=10)     466.00 ±  0%  +870.83% (p=0.000 n=10)
Processor/SetAttributesBatch-16     648.0 ± 3%     1089.5 ± 1%   +68.13% (p=0.000 n=10)     1087.5 ±  4%   +67.82% (p=0.000 n=10)
LoggerNewRecord/5_attributes-16     0.000 ± 0%      0.000 ± 0%         ~ (p=1.000 n=10) ¹    0.000 ±  0%         ~ (p=1.000 n=10) ¹
LoggerNewRecord/10_attributes-16    610.0 ± 0%      610.0 ± 0%         ~ (p=1.000 n=10) ¹    610.0 ±  0%         ~ (p=1.000 n=10) ¹
LoggerProviderLogger-16             354.5 ± 6%      368.0 ± 7%         ~ (p=0.288 n=10)      391.0 ± 29%         ~ (p=0.239 n=10)
WalkAttributes/1_attributes-16      0.000 ± 0%      0.000 ± 0%         ~ (p=1.000 n=10) ¹    0.000 ±  0%         ~ (p=1.000 n=10) ¹
WalkAttributes/10_attributes-16     0.000 ± 0%      0.000 ± 0%         ~ (p=1.000 n=10) ¹    0.000 ±  0%         ~ (p=1.000 n=10) ¹
WalkAttributes/100_attributes-16    0.000 ± 0%      0.000 ± 0%         ~ (p=1.000 n=10) ¹    0.000 ±  0%         ~ (p=1.000 n=10) ¹
WalkAttributes/1000_attributes-16   0.000 ± 0%      0.000 ± 0%         ~ (p=1.000 n=10) ¹    0.000 ±  0%         ~ (p=1.000 n=10) ¹
SetAddAttributes/SetAttributes-16   48.00 ± 0%      48.00 ± 0%         ~ (p=1.000 n=10) ¹    48.00 ±  0%         ~ (p=1.000 n=10) ¹
SetAddAttributes/AddAttributes-16   0.000 ± 0%      0.000 ± 0%         ~ (p=1.000 n=10) ¹    0.000 ±  0%         ~ (p=1.000 n=10) ¹
SimpleProcessorOnEmit-16            0.000 ± 0%      0.000 ± 0%         ~ (p=1.000 n=10) ¹    0.000 ±  0%         ~ (p=1.000 n=10) ¹
geomean                                        ²                ?                       ²                 ?                       ²
¹ all samples are equal
² summaries must be >0 to compute geomean

                                  │   old.txt    │                new.txt                │              new-pgo.txt              │
                                  │  allocs/op   │ allocs/op   vs base                   │ allocs/op   vs base                   │
BatchProcessorOnEmit-16             0.000 ± 0%     0.000 ± 0%         ~ (p=1.000 n=10) ¹   0.000 ± 0%         ~ (p=1.000 n=10) ¹
Processor/Simple-16                 0.000 ± 0%     1.000 ± 0%         ? (p=0.000 n=10)     1.000 ± 0%         ? (p=0.000 n=10)
Processor/Batch-16                  0.000 ± 0%     1.000 ± 0%         ? (p=0.000 n=10)     1.000 ± 0%         ? (p=0.000 n=10)
Processor/SetTimestampSimple-16     0.000 ± 0%     1.000 ± 0%         ? (p=0.000 n=10)     1.000 ± 0%         ? (p=0.000 n=10)
Processor/SetTimestampBatch-16      0.000 ± 0%     1.000 ± 0%         ? (p=0.000 n=10)     1.000 ± 0%         ? (p=0.000 n=10)
Processor/AddAttributesSimple-16    0.000 ± 0%     1.000 ± 0%         ? (p=0.000 n=10)     1.000 ± 0%         ? (p=0.000 n=10)
Processor/AddAttributesBatch-16     0.000 ± 0%     1.000 ± 0%         ? (p=0.000 n=10)     1.000 ± 0%         ? (p=0.000 n=10)
Processor/SetAttributesSimple-16    1.000 ± 0%     2.000 ± 0%  +100.00% (p=0.000 n=10)     2.000 ± 0%  +100.00% (p=0.000 n=10)
Processor/SetAttributesBatch-16     1.000 ± 0%     2.000 ± 0%  +100.00% (p=0.000 n=10)     2.000 ± 0%  +100.00% (p=0.000 n=10)
LoggerNewRecord/5_attributes-16     0.000 ± 0%     0.000 ± 0%         ~ (p=1.000 n=10) ¹   0.000 ± 0%         ~ (p=1.000 n=10) ¹
LoggerNewRecord/10_attributes-16    4.000 ± 0%     4.000 ± 0%         ~ (p=1.000 n=10) ¹   4.000 ± 0%         ~ (p=1.000 n=10) ¹
LoggerProviderLogger-16             1.000 ± 0%     1.000 ± 0%         ~ (p=1.000 n=10) ¹   1.000 ± 0%         ~ (p=1.000 n=10) ¹
WalkAttributes/1_attributes-16      0.000 ± 0%     0.000 ± 0%         ~ (p=1.000 n=10) ¹   0.000 ± 0%         ~ (p=1.000 n=10) ¹
WalkAttributes/10_attributes-16     0.000 ± 0%     0.000 ± 0%         ~ (p=1.000 n=10) ¹   0.000 ± 0%         ~ (p=1.000 n=10) ¹
WalkAttributes/100_attributes-16    0.000 ± 0%     0.000 ± 0%         ~ (p=1.000 n=10) ¹   0.000 ± 0%         ~ (p=1.000 n=10) ¹
WalkAttributes/1000_attributes-16   0.000 ± 0%     0.000 ± 0%         ~ (p=1.000 n=10) ¹   0.000 ± 0%         ~ (p=1.000 n=10) ¹
SetAddAttributes/SetAttributes-16   1.000 ± 0%     1.000 ± 0%         ~ (p=1.000 n=10) ¹   1.000 ± 0%         ~ (p=1.000 n=10) ¹
SetAddAttributes/AddAttributes-16   0.000 ± 0%     0.000 ± 0%         ~ (p=1.000 n=10) ¹   0.000 ± 0%         ~ (p=1.000 n=10) ¹
SimpleProcessorOnEmit-16            0.000 ± 0%     0.000 ± 0%         ~ (p=1.000 n=10) ¹   0.000 ± 0%         ~ (p=1.000 n=10) ¹
geomean                                        ²               ?                       ²               ?                       ²
¹ all samples are equal
² summaries must be >0 to compute geomean
```
2024-08-01 10:13:43 +02:00
Robert Pająk
d595dab76f sdk/log: Simple processor may be useful for production (#5578)
Update the design doc:
- describe that minimizing the amount of the heap allocation is one of
the design goals
- provide more usage examples

Moreover update the docs to not discourage using a simple processor from
production. It would be good if the user would use it e.g. with an
[user_events](https://docs.kernel.org/trace/user_events.html),
[LLTng](https://lttng.org/docs/v2.13/#doc-tracing-your-own-user-application)
or
[ETW](https://learn.microsoft.com/en-us/windows/win32/etw/about-event-tracing)
exporter. However, we recommend using a batch processor when using OTLP
over HTTP or gRPC exporters.
2024-07-11 20:06:19 +02:00
Robert Pająk
6082e830c2 sdk/log: Remove slice allocation from SimpleProcessor.OnEmit (#5493)
The reason for this improvement (apart that, in general, it is good to
have better performance) is there may be good use case to use a
`SimpleProcessor` to emit logs efficiently to standard output. It could
be one of the most efficient solutions (from application performance
perspective) and thanks to such configuration the user would not lose
any logs if the application suddenly crashes. For instance, a useful
configuration could be a simple processor with an OTLP file exporter
(https://github.com/open-telemetry/opentelemetry-go/issues/5408).

I think we might consider changing the following portion of
`SimpleProcessor` documentation (but I would prefer to do it as separate
PR):

> // This Processor is not recommended for production use. The
synchronous
// nature of this Processor make it good for testing, debugging, or
// showing examples of other features, but it can be slow and have a
high
// computation resource usage overhead. [NewBatchProcessor] is
recommended
// for production use instead.

```
goos: linux
goarch: amd64
pkg: go.opentelemetry.io/otel/sdk/log
cpu: Intel(R) Core(TM) i9-10885H CPU @ 2.40GHz
                                    │   old.txt   │               new.txt                │
                                    │   sec/op    │    sec/op     vs base                │
Processor/Simple-16                   449.4n ± 7%   156.2n ±  5%  -65.25% (p=0.000 n=10)
Processor/ModifyTimestampSimple-16    468.0n ± 6%   171.3n ± 15%  -63.40% (p=0.000 n=10)
Processor/ModifyAttributesSimple-16   515.8n ± 3%   233.2n ±  8%  -54.77% (p=0.000 n=10)
geomean                               476.9n        184.1n        -61.40%

                                    │   old.txt   │                 new.txt                 │
                                    │    B/op     │    B/op     vs base                     │
Processor/Simple-16                    417.0 ± 0%     0.0 ± 0%  -100.00% (p=0.000 n=10)
Processor/ModifyTimestampSimple-16     417.0 ± 0%     0.0 ± 0%  -100.00% (p=0.000 n=10)
Processor/ModifyAttributesSimple-16   465.00 ± 0%   48.00 ± 0%   -89.68% (p=0.000 n=10)
geomean                                432.4                    ?                       ¹ ²
¹ summaries must be >0 to compute geomean
² ratios must be >0 to compute geomean

                                    │  old.txt   │                 new.txt                 │
                                    │ allocs/op  │ allocs/op   vs base                     │
Processor/Simple-16                   1.000 ± 0%   0.000 ± 0%  -100.00% (p=0.000 n=10)
Processor/ModifyTimestampSimple-16    1.000 ± 0%   0.000 ± 0%  -100.00% (p=0.000 n=10)
Processor/ModifyAttributesSimple-16   2.000 ± 0%   1.000 ± 0%   -50.00% (p=0.000 n=10)
geomean                               1.260                    ?                       ¹ ²
¹ summaries must be >0 to compute geomean
² ratios must be >0 to compute geomean
```

---------

Co-authored-by: Tyler Yahn <MrAlias@users.noreply.github.com>
2024-06-11 07:28:56 -07:00
Robert Pająk
ed666f7713 sdk/log: Rename BatchingProcessor to BatchProcessor (#5229) 2024-04-19 08:34:30 +02:00
Robert Pająk
054a63fef2 sdk/log: Add options to NewSimpleProcessor (#5185) 2024-04-10 19:47:40 +02:00
Robert Pająk
b9752eb5dc sdk/log: Exporter has to be concurrent-safe (#5181) 2024-04-09 10:22:50 +02:00
Tyler Yahn
12c5651ec7 Do not alloc multiple noop exporters (#5083) 2024-03-19 10:33:57 -07:00
Tyler Yahn
3542ee68a9 Implement the SimpleProcessor (#5079)
* Implement the SimpleProcessor

* Add BenchmarkSimpleProcessorOnEmit

* Remove stale comment

* Run benchmarks in parallel
2024-03-17 07:47:05 -07:00
Tyler Yahn
da047e70ef Add the Enabled method to the Logger (#5071)
* Add the Enabled method to the Logger

* Add a changelog entry

* Rename enabled.go to min_sev.go

* Remove MinSeverityProcessor

* Document lack of interaction between OnEmit and Enabled

* Update sdk/log/processor.go

---------

Co-authored-by: Robert Pająk <pellared@hotmail.com>
2024-03-15 08:15:44 -07:00
Robert Pająk
54b6ee4174 sdk/log: Scaffolding (#5068) 2024-03-13 17:47:07 +01:00