Attender: Guohao, Minyu, Yiyang, XiChen, Hongbing, Guangbao, Yuanben, Conway Zhang, Sammi Agenda:
1. Shopee - 1.4.0 Release RATIS 3.0.0 enabling in Ozone is ongoing(HDDS-9192 <https://issues.apache.org/jira/browse/HDDS-9192>). There is one remaining OM RATIS snapshot test failure to address. In the meanwhile, RATIS community has started the discussion of 3.0.0 release. William from IotDB will be the RM. - Some remaining block deletion improvement JIRAs need review. HDDS-8882 <https://issues.apache.org/jira/browse/HDDS-8882>, HDDS-8888 <https://issues.apache.org/jira/browse/HDDS-8888> 1. DiDi - Deploying a new cluster with the master branch. - One existing cluster, the DN storage utilization is very uneven due to new nodes added later. After disabled the compression in container replication and tuned some parameters of Container Balancer, the speed of container move cannot catch up with the speed of data ingestion. A storage utilization considered pipeline choose policy is proposed, HDDS-9345 <https://issues.apache.org/jira/browse/HDDS-9345> and HDDS-9449 <https://issues.apache.org/jira/browse/HDDS-9449>. - In one existing cluster(1.3.0), it's found that there are block deletion transaction records in SCM DB, which have different tx IDs, but the same set the blocks to delete. The root cause is not found yet. - Proposed a solution to reduce the RPC call from client to OM, to improve the performance. HDDS-9448 <https://issues.apache.org/jira/browse/HDDS-9448>. 1. Qihoo - Deployed new clusters with master branch recently. The clusters are running well so far. - Currently focused on end-to-end latency glitch optimization, key targets are P95 and P99 latency. Has found and proposed two optimizations, HDDS-9330 <https://issues.apache.org/jira/browse/HDDS-9330> block allocation performance, and HDDS-9417 <https://issues.apache.org/jira/browse/HDDS-9417> OM key deletion lock optimization.
