You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@pulsar.apache.org by Apache Pulsar Slack <ap...@gmail.com> on 2020/05/28 09:11:03 UTC

Slack digest for #dev - 2020-05-28

2020-05-27 18:50:18 UTC - Addison Higham: so small update to this issue ^^ changing os level keep alive does fix the issue. When a bookie gets rescheduled, the time for a broker to be happy is just how fast keep alive.

 On k8s (or anywhere really), it is a bit of pain to mess with those settings so would still be nice to make that an app level thing.

Sounds like the best path would be to implement a heartbeat in the connections to bookkeeper and to kill sockets if heartbeats fail
----
2020-05-27 22:09:06 UTC - Kevin Xu: @Kevin Xu has joined the channel
----
2020-05-28 08:33:57 UTC - Sunny Chan: @Sunny Chan has joined the channel
----