Thanks for bumping this. I do have a concern:

This proposal changes the names of existing metrics - as such, it will
require all owners of monitoring systems to update their dashboards. It
will also complicate monitoring of multiple clusters with different
versions and require some modifications to existing monitoring automation

What do you think of an alternative solution:
1. For the next release, add the validations on the broker side and print a
"warning" that this client id is invalid, that it will break metrics and
that it will be rejected in newer versions.
2. Few releases later, actually turn the validation on and return
InvalidClientID error to clients.

We did something similar when we deprecated acks=2.

On Thu, Aug 31, 2017 at 12:13 PM Mickael Maison <[EMAIL PROTECTED]>
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB