Log spam about timeout not being set on GetTabletListToPollForCDC

[ccorbacho]

• Release: v1.9.5.y.25

We made a change today to reduce the cdc.poll.interval from the default 500ms to 10ms, and we keep seeing these multiple attempts and log lines for the same tablet in the Kafka Connect logs:

{"@timestamp":"2023-07-12T10:07:24.883Z","source_host":"<our kafka connector>","file":"TabletClient.java","method":"sendRpc","level":"WARN","line_number":"158","thread_name":"pool-134-thread-1","@version":1,"logger_name":"org.yb.client.TabletClient","message":"[Peer a9140d770a004e61864b52fc74df0efb]  sending an rpc without a timeout YRpc(method=GetTabletListToPollForCDC, service=yb.cdc.CDCService, tablet=0e324cbc59a74f5dbc01a6545db90a76, attempt=1, maxAttempts=100, maxTimeoutMs=0, elapsedTimeMs=518)","class":"org.yb.client.TabletClient","mdc":{}}

{"@timestamp":"2023-07-12T10:07:24.926Z","source_host":"<our kafka connector>","file":"TabletClient.java","method":"sendRpc","level":"WARN","line_number":"158","thread_name":"yb-nio-1","@version":1,"logger_name":"org.yb.client.TabletClient","message":"[Peer f31e98ae09c54836a65fa9a4e46283c6]  sending an rpc without a timeout YRpc(method=GetTabletListToPollForCDC, service=yb.cdc.CDCService, tablet=0e324cbc59a74f5dbc01a6545db90a76, attempt=2, maxAttempts=100, maxTimeoutMs=0, elapsedTimeMs=1037)","class":"org.yb.client.TabletClient","mdc":{}}

From a quick look over the connector and the client code, I cannot see why it is complaining about the timeout not being set - we are not overriding the timeout being used here (which, AFAICT, would be the admin.operation.timeout.ms default for GetTabletListToPollForCDC), and the code looks like it should be setting this correctly.