Re: [DISCUSS] Doubt about the widely used Netty Recycler in Pulsar

2025-02-22 Thread Enrico Olivelli
Yunze, I support your doubts. Recycling objects is useful only under certain conditions and it adds lot of complexity. It is a tool that we can use but we should do it only when needed. And 'when needed' can only be measured with running workloads that exercise the code paths that you are changing

Re: [VOTE] Release Apache Pulsar 3.3.5 based on 3.3.5-candidate-1

2025-02-22 Thread Yike Xiao
+1 (non-binding) - Built from source - Checked the signatures of the source and binary release artifacts - Ran pulsar standalone - Checked producer and consumer - Verified the Cassandra connector - Verified the Stateful function Regards, Yike From: Lari Hotari S

Re: [VOTE] Release Apache Pulsar 4.0.3 based on 4.0.3-candidate-1

2025-02-22 Thread Yike Xiao
+1 (non-binding) - Built from source - Checked the signatures of the source and binary release artifacts - Ran pulsar standalone - Checked producer and consumer - Verified the Cassandra connector - Verified the Stateful function Regards, Yike From: Lari Hotari S

Re: [VOTE] Release Apache Pulsar 3.0.10 based on 3.0.10-candidate-1

2025-02-22 Thread Yike Xiao
+1 (non-binding) - Built from source - Checked the signatures of the source and binary release artifacts - Ran pulsar standalone - Checked producer and consumer - Verified the Cassandra connector - Verified the Stateful function Regards, Yike From: Lari Hotari S

Re: [DISCUSS] Doubt about the widely used Netty Recycler in Pulsar

2025-02-22 Thread Yunze Xu
Bump this thread again. I really suffered from the use of Netty recyclers. It makes code much really harder to maintain. I also made a benchmark and the recycler allocation is 10~20x slower than a normal heap allocation. https://gist.github.com/BewareMyPower/3dcc59183c92c76e9c985cced16d049d I kno