[Release/7.0] Don't emit manifest for NativeRuntimeEventSource (#78213) #79009
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Outdated suggestions cannot be applied.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.
Port #78213 to 7.0
Customer Impact
When starting an ETW session on a machine with lots of .net processes every 6.0 and newer process will emit a very large manifest event for NativeRuntimeEventSource (~50kb per process). This causes an event storm and will overflow buffers leading to dropped events. See #77014 for a customer impacted.
Historically we have never emitted the manifest for the
Microsoft-Windows-DotNETRuntime*
providers, consumers need to hardcode each event, or use a library such as TraceEvent that parses it for you. In 6.0 we added the ability for managed BCL code to emit events from theMicrosoft-Windows-DotNETRuntime
provider by adding events to NativeRuntimeEventSource, previously it would not have any events declared and only existed to forward events to in process EventListeners. Making NativeRuntimeEventSource have events unintentionally caused us to start emitting the manifest.Regression?
Yes
Testing
Manual testing that the manifest is not emitted with this change
Risk
Low