#influxdb

is there a query/something that would show me memory usage by resources/queries?

our influx server is taking all the memory that i give him

this is not cool

i double the size of machine... and again... i almoust ran out of memory?!?!

i know all about mmaped data, etc. But my data in cca 5 databases is on filesystem... below 1gb.

and i take around 22 gb of ram already

the problem is that i don't know how to efficiently examine the influx. so that's why i am asking for queries that would show me more on memory usage so i could dig dipper to solve this

hello

i have an issue with procstat plugin

dors somebody can help me

telegraf run as telegraf service

excuse me telegraf service run as telegraf user

so when i configure a service like this /etc/telegraf/telegraf.d/bapplipuppet-procstat.conf [[inputs.procstat]] prefix = "bapplipuppet" user = "bapplipuppet"

telegraf reports me 2016/09/02 17:36:00 Error: procstat getting process, exe: [] pidfile: [] pattern: [] user: [bapplipuppet] Failed to execute /usr/bin/pgrep. Error: 'exit status 1'

i have a workaround to run telegraf as root

But i would like to not use this workaround if not necessary.

Does somebody have a trick to do it?

Hey All. I think I hit a bug in influxdb. I was logging haproxy data. Had maybe 3 days of data in there, and I did a: select * from haproxy limit 10

It ate through 8gb of ram, and around 14gb of swap before I killed it to prevent it from OOM'ing.