apache / datafusion-ballista

Apache DataFusion Ballista Distributed Query Engine
https://datafusion.apache.org/ballista
Apache License 2.0
1.42k stars 183 forks source link

The last_seen_ts_threshold for getting alive executor at the scheduler side should be larger than the heartbeat time interval #785

Closed yahoNanJing closed 1 year ago

yahoNanJing commented 1 year ago

Describe the bug

Currently the last_seen_ts_threshold for getting alive executor at the scheduler side is one minute. While the heartbeat between executor and scheduler is also 60 seconds. It will occasionally happen that alive executors fail to be detected due to hearbeat too late.

To Reproduce

Expected behavior

Additional context