Details
-
Sub-task
-
Status: Open
-
Major
-
Resolution: Unresolved
-
None
-
None
-
None
Description
In reviewing https://github.com/apache/kafka/pull/3987 I noticed that the autogenerated tables currently differ from the manually created ones. The manual ones have 3 columns – metric/attribute name, description, and mbean name with a regex. The new ones have 3 columns, but for some reason the first one is just empty, the second is the metric/attribute name, the last one is the description, and there is no regex.
We could potentially just drop to two columns since the regex column is generally very repetitive and is now handled by a header row giving the general group mbean name info/format. The one thing that seems to currently be missing is the regex that would restrict the format of these (although these weren't really technically enforced and some of the restrictions are being removed, e.g. see some of the follow up discussion to https://cwiki.apache.org/confluence/display/KAFKA/KIP-190%3A+Handle+client-ids+consistently+between+clients+and+brokers).