Hi Siddharth,
Thank you,
We're using version 2.2.0.0 with about 150 hosts.
Did not find any error like the one we have on confluence.
I've set timeline.metrics.cluster.aggregator.second.ttl from 15 days to 3,
will see if that helps.
Regards,
Olivier
On Wed, Sep 21, 2016 at 6:04 PM, Siddharth
Hi Eric,
Any reason for using 2.2.0 ? There we a bunch of issues fixed in the later
versions with regards to AMS. I would suggest using at least 2.2.2 for
performance reasons.
Cluster level data is much lesser in volume than Host level.
So in terms of ttl,
*.host.aggregator.* makes more sen