Github user mgummelt commented on the issue:
https://github.com/apache/spark/pull/17750
1. Please don't add any features to fine-grained mode. It's deprecated.
2. The linked JIRA, as well as the title for this PR, only refer to
checkpointing, not failover timeouts, which are a different feature. These two
features should be separate JIRAs and separate PRs.
3. As @lins05 has explained, the Spark driver is not designed to recover
from a crash. Recovery is not as simple as adding a failover timeout.
I am looking at solving a problem where an intermittent network partition
can result in the driver being killed unnecessarily, and it's possible that
adding a failover_timeout will solve that, but I'm still looking into that.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
