If the difference between the record createTime timestamp and logAppendtime timestamp is high, how can we tell whether the broker is slow or the producer?
“request-latency-avg” - The avg request latency in ms
“record-queue-time-avg” - The average time in ms record batches spent in the send buffer.
What is that single metric or combination of metrics of producer or broker that can help us investigate which side we need to investigate, Producer or Broker
onprem or cloud deployment?
the following might be helpful
k8s or something similar or tarball/package installation?
trying to get the bigger picture