You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@pulsar.apache.org by "thetumbled (via GitHub)" <gi...@apache.org> on 2024/07/16 09:08:41 UTC
[I] [Transaction] One single TC unavailable cause transactional throughput down to 0. [pulsar]
thetumbled opened a new issue, #23039:
URL: https://github.com/apache/pulsar/issues/23039
### Search before asking
- [X] I searched in the [issues](https://github.com/apache/pulsar/issues) and found nothing similar.
### Motivation
Transaction module introduce multiple(default 16) TC to improve performance.
However, if part of TC fail to recover, the throughput of all transactional client down to 0, though there are some TC able to serve for client.
<img width="646" alt="image" src="https://github.com/user-attachments/assets/c1ae74f7-49e7-4fa0-9f49-7c3fa8db5072">
### Solution
When there are some TC able to provide service, transactional client should seek service from these TC, instead of waiting for failed TC recovery.
### Alternatives
_No response_
### Anything else?
_No response_
### Are you willing to submit a PR?
- [X] I'm willing to submit a PR!
--
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: commits-unsubscribe@pulsar.apache.org.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [I] [Transaction] One single TC unavailable cause transactional throughput down to 0. [pulsar]
Posted by "thetumbled (via GitHub)" <gi...@apache.org>.
thetumbled commented on issue #23039:
URL: https://github.com/apache/pulsar/issues/23039#issuecomment-2230421267
Discuss link: https://lists.apache.org/thread/w965ygdgmn6yjpkp3lx23d3lo999tt88
--
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: commits-unsubscribe@pulsar.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
Re: [I] [Transaction] One single TC unavailable cause transactional throughput down to 0. [pulsar]
Posted by "merlimat (via GitHub)" <gi...@apache.org>.
merlimat commented on issue #23039:
URL: https://github.com/apache/pulsar/issues/23039#issuecomment-6388755761
Partly addressed: with the 5.0 metadata-store coordinator election (PIP-473, #25929), `WatchTcAssignmentsDiscovery.nextHandler()` skips coordinators that have no elected leader. A coordinator that has a leader but fails to recover is still picked round-robin, and so is any unavailable coordinator under the older `AssignTopicTcDiscovery`. Making `nextHandler()` skip handlers that aren't connected and ready, or that just failed `newTransaction`, would fix the throughput drop. This is client-internal and doesn't need a protocol change.
--
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: commits-unsubscribe@pulsar.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org