Zhihong Yu

I saw the following in kubelet log: ``` Jun 12 13:36:27 centos-linux.shared kubelet[1253]: E0612 13:36:27.015557 1253 remote_runtime.go:200] CreateContainer in sandbox "c6f5403980fda62601142e939a67042ed8e173b829b52e309db84ecd1d3f27ae" from runtime service failed: rpc error: code = Unknown...

For kubelet running on containerd, 'pod_workers.go:191' ("failed to create containerd task) occurred once. For kubelet running on cri-o, 'pod_workers.go:191' occurred several times.

Can the checkstyle warnings for hbase-hadoop-compat be addressed ?

Thanks for the PR. I am a bit curious: previously there were incrSplitRequest and the success count. In the PR, there is also failure count which can be deduced from...

Yes, please.

+1 Can rerun tests to see if they pass.

Which JUnit release contains this change ? Thanks

cc @sunchao

If I subclass `UnknownPartitioning` directly, I would get this compilation error: ``` [error] /nfusr/dev-server/zyu/spark-cassandra-connector/connector/src/main/scala/com/datastax/spark/connector/datasource/CassandraScanBuilder.scala:327:92: not enough arguments for constructor UnknownPartitioning: (x$1: Int)org.apache.spark.sql.connector.read.partitioning.UnknownPartitioning. [error] Unspecified value parameter x$1. [error] case class...

If custom partitioning reports `UnknownPartitioning` to Spark and can keep 3.2.1 behavior, that means the current check is not desired.