Memory leaks in post-process-forwarder-transactions

[doc-sheet]

Self-Hosted Version

24.10.0

CPU Architecture

x86_64

Docker Version

20.10.17

Docker Compose Version

none

Steps to Reproduce

1. Run post-process-forwarder-transactions with multiple processes

sentry run consumer --no-strict-offset-reset post-process-forwarder-transactions --consumer-group=post-process-forwarder --synchronize-commit-log-topic=snuba-transactions-commit-log --synchronize-commit-group=transactions_group -- --processes 4 --concurrency 4

2. Wait

Expected Result

Normal work just as any other consumer

Actual Result

Consumer being stopped by oom-killer.

I tried different limits - but if only delays inevitable OOM.

There are no logs in between "Starting multithreaded post process forwarder" and crash.

Event ID

No response

[doc-sheet]

Sorry if wrong project, I just unsure if issue in my setup or sentry itself.

[getsantry]

Routing to @getsentry/product-owners-ingestion-and-filtering for triage ⏲️

[getsantry]

Assigning to @getsentry/support for routing ⏲️

[aldy505]

Nothing we can do in self-hosted. Should be transferred to sentry for triage. cc @bc-sentry