apache / apache/beam

Don't leak garbage collection timers in GlobalWindow

Open
#17,988 0 comments 0 reactions 0 assignees View on GitHub
bug core P3 runners
Dominant language
Java
Stars
8.7k
Forks
4.7k
Avg merge
1d 20h
Merged PRs (30d)
196

Description

Consider the transform:
Window
.into(new GlobalWindows())
.triggering(
Repeatedly.forever(
AfterProcessingTime.pastFirstElementInPane().plusDelayOf(...)))
.discardingFiredPanes()
This is a common idiom for 'process elements bunched by arrival time'.
Currently we create an end-of-window timer per key, which clearly will only fire if the pipeline is drained.
Better would be to avoid creating end-of-window timers if there's no state which needs to be processed at end-of-window (ie at drain if the Global window).

Imported from Jira [BEAM-175](https://issues.apache.org/jira/browse/BEAM-175). Original Jira may contain additional context.
Reported by: mshields822.

Contributor guide

Open the contributing guide

Research direction

The issue names GlobalWindow/GlobalWindows, Repeatedly, AfterProcessingTime, and end-of-window timers, but no source file or test. Start by tracing how these windowing and trigger components create timers for each key, then verify the arrival-time pattern does not retain unnecessary end-of-window timers while still handling state that must be processed at window end.

Written by the indexing model from the issue text.

Assessment

Tech stack
java
Domain
data-engineering, distributed-systems
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.