Github user arunmahadevan commented on the pull request:
https://github.com/apache/storm/pull/644#issuecomment-128266209
I was trying to put a safety net so that the recovery would always work.
The limitation on the timed rotation policy is also on similar lines that we
don't know how much data would be written to the files before the timed
rotation kicks in.
So,
- For file size rotation, remove the restrictions on the file size and just
log warnings.
- For time based rotation, set a flag and do the actual rotation in
`doCommit` as you suggested.
- Add a note in the README about the risk that the recovery would fail if
files cannot be recovered within timeout and hence should be kept to reasonable
sizes (and time interval) or the message timeout should be increased.
Does it sound reasonable ? If we agree, will make the changes.
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
