junaiddshaukat opened a new issue, #37465:
URL: https://github.com/apache/beam/issues/37465
### What happened?
### What needs to happen?
The `CombinePerKeyPrecombineOperator` in the TypeScript SDK currently uses a
suboptimal cache eviction strategy. When the cache exceeds `maxKeys`, it
flushes entries in insertion order rather than based on actual usage patterns.
**Current behavior (lines 485-488 in operators.ts):**
if (this.groups.size > this.maxKeys) {
// Flush a random 10% of the map to make more room.
// TODO: Tune this, or better use LRU or ARC for this cache.
return this.flush(this.maxKeys * 0.9);
}
### Proposed change:
Implement LRU (Least Recently Used) cache eviction, similar to what was done
for CachingStateProvider in PR #37214. This ensures that frequently accessed
keys are kept in cache while infrequently used keys are evicted first.
### Issue Priority
Priority: 2 (default / most bugs should be filed as P2)
### Issue Components
- [ ] Component: Python SDK
- [ ] Component: Java SDK
- [ ] Component: Go SDK
- [x] Component: Typescript SDK
- [ ] Component: IO connector
- [ ] Component: Beam YAML
- [ ] Component: Beam examples
- [ ] Component: Beam playground
- [ ] Component: Beam katas
- [ ] Component: Website
- [ ] Component: Infrastructure
- [ ] Component: Spark Runner
- [ ] Component: Flink Runner
- [ ] Component: Samza Runner
- [ ] Component: Twister2 Runner
- [ ] Component: Hazelcast Jet Runner
- [ ] Component: Google Cloud Dataflow Runner
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
