#influxdb

ello a stupid question, I have setup influxdb and telegraf everything working very well, I have a a ceph-cluster and wondering if I should enable ceph plugins on monitor-node or on the OSD?

last week i was asking about memory leaks, problematic queries, etc. During this weekend my influxdb crashed since memory run out.

I switched from 14gb to 28gb now to have more space to find the problem

but i still haven't find a way/tools to deal with this kind of problems.

isn't there a way to see for instance queries which takes/consumes most memory?

i was told that data is mmaped in influxdb which is ok, but still i can't be this, since the whole data directory is 2Gb?!?!

and it crashed the system with 14gb

i have 7 dbs, 12 continuous queries, nothing special really

i would appreciate any help/link to blog/tutorial on performance/anything which would help me solved this

anybody?

hmm to allow telegraf to monitoring ceph must I add the user ti ceph group or?

can you limit memory for influxdb?