Tracing aware executors wrap Runnables and Callables more cleanly #310
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Generate changelog in
|
|
@carterkozak I actually think your description is great as a commit message instead of one sentence summary. |
|
Good call, updated |
|
Is there a way of including an assertion about the stackframes here? The new ones are very neat and don't want to accidentally regress! |
|
Sure, I did something similar in tritium: https://github.com/palantir/tritium/blob/d69b494360ed69a0739345124028e16e52029821/tritium-lib/src/test/java/com/palantir/tritium/proxy/InstrumentationTest.java#L555-L569 will put it together shortly |
Firstly, we should not optimize around stack depth in most cases
because it can result in large methods that the JVM cannot efficiently
inline and additional GC pressure depending on the implementation.
However, we use tracing a lot, a tracing wrapped executor shows up
in ~50% of stack traces in a large internal service, and exceptions
propagate through multiple executors in many cases.
This would not make sense to change if the tracing stack frames
provided information, but wrapping from runnable to callable
isn't helpful, nor are lambda frames to turn a Callable into
a ThrowingCallable.
This change is completely internal to tracing and does not expose
any new public APIs.
Before: frames 2, 3, 4, and 5 are from tracing
```
com.palantir.logsafe.exceptions.SafeRuntimeException: Stack Trace
at com.palantir.tracing.TracerTest.lambda$testStackDepth$9(TracerTest.java:486)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
at com.palantir.tracing.DeferredTracer.withTrace(DeferredTracer.java:104)
at com.palantir.tracing.Tracers$TracingAwareCallable.call(Tracers.java:356)
at com.palantir.tracing.WrappingExecutorService.lambda$wrapTask$0(WrappingExecutorService.java:67)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
```
After: line 2 handles tracing, the rest would be present without tracing
```
com.palantir.logsafe.exceptions.SafeRuntimeException: Stack Trace
at com.palantir.tracing.TracerTest.lambda$testStackDepth$9(TracerTest.java:486)
at com.palantir.tracing.Tracers$TracingAwareRunnable.run(Tracers.java:409)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
```
carterkozak
force-pushed
the
ckozak/tracing_stack_frames
branch
from
bc57f2d to
5fa5616
Compare
|
Updated, added docs to the tests to make the intent clear. |
|
Released 3.3.0 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
==COMMIT_MSG==
Firstly, we should not optimize around stack depth in most cases
because it can result in large methods that the JVM cannot efficiently
inline and additional GC pressure depending on the implementation.
However, we use tracing a lot, a tracing wrapped executor shows up
in ~50% of stack traces in a large internal service, and exceptions
propagate through multiple executors in many cases.
This would not make sense to change if the tracing stack frames
provided information, but wrapping from runnable to callable
isn't helpful, nor are lambda frames to turn a Callable into
a ThrowingCallable.
This change is completely internal to tracing and does not expose
any new public APIs.
Before: frames 2, 3, 4, and 5 are from tracing
After: line 2 handles tracing, the rest would be present without tracing
==COMMIT_MSG==
Possible downsides?
none.