that looks good - ideally when a sink or source fail they should gracefully die and restart without any loss of data (or gracefully handle it in some way)

or something like that - nice on Yike but could we get an official stance on this issue? Perhaps the pulsar team already have an approach around this?

Hello community Pulsar, Can the sateful functions be activated on a pulsar 3.0.0 in standalone mode ?

I presented this PIP and its background in Pulsar Summit Europe 2023 if you fancy a video instead of reading the PIP.

This message was deleted.

This message was deleted.

This message was deleted.

Good afternoon to all

This message was deleted.

This message was deleted.

The Apache Pulsar community meets every 2 weeks virtually. We discuss open Pull Requests, Proposals (PIP), future roadmaps and just chit chatting on everything Pulsar. Join us! It’s today at 15:30 UTC All details are here: https://github.com/apache/pulsar/wiki/Community-Meetings

This message was deleted.

This message was deleted.

This message was deleted.

hey team - another question. Gone ahead to use HPA for pod instances (for some aspects of our pipelines) but encountered a potential issue - I understand it is 1 instance per pod i.e one handler, and 1 message ingested at a time. But when the scale reaches what we expect i.e 500events/second - the CPU and memory usage go up. Why does this occur, as i thought each pod was single threaded, why would they need additional memory & cpu usage if they are built to only handle 1 message at a time. I think i am not understanding something, so would appreciate guidance

This message was deleted.

This message was deleted.

This message was deleted.

This message was deleted.

This message was deleted.

This message was deleted.

This message was deleted.

This message was deleted.

I wonder how netty’s support for IO_uring is these days.

Hi, I'm trying to update the HomeBrew

apache-pulsar

formula to 3.0.0 https://github.com/Homebrew/homebrew-core/pull/133460/files, Pulsar cli arguments seem to have changed since 2.10.3 2.10.3 standalone was using

bin/"pulsar", "standalone", "--zookeeper-dir", "#{testpath}/zk", " --bookkeeper-dir", "#{testpath}/bk"

and for tests

#{bin}/pulsar initialize-cluster-metadata -c a -cs localhost -uw localhost -zk localhost

I've changed for 3.0.0 standalone

bin/"pulsar", "standalone", "--metadata-dir", "#{testpath}/metadata", " --bookkeeper-dir", "#{testpath}/bk"

but the

initialize-cluster-metadata

is failing https://github.com/Homebrew/homebrew-core/actions/runs/5265568983/jobs/9518538264?pr=133460 Any ideas?

This message was deleted.

This message was deleted.

https://github.com/apache/pulsar/discussions/20311 cherry-pick this usage issue that I'm interested in also. Perhaps anyone had similar use case can share the best practices?

Hello All, I have setup pulsar on kubernetes on the latest version 2.11 and after restarting pulsar-broker. I am getting the below error

broker pulsar-broker-0.pulsar-broker.pulsar.svc.cluster.local znode still exists ... check in 10 seconds ...