Graylog.cluster_config high latency in Production stage

Hi guys,

Currently, my mongodb of graylog production stage reached 99% of CPU Usage.
This happened suddenly in the afternoon.

Still figuring out the root cause but i have no idea how to debug deeply.

I have scaled up the mongodb VM from 2 cores to 4 cores but still reach ~99%

My Graylog version is

Version:
3.2.6+e255fcc, codename *Ethereal

Wondering if you guys can share or have experienced more in mongodb and help me to solve this. Really appreciate :pray:

Hi Merces,

Not running a cluster here so not really a clue.

You could start at the server logfile of graylog and see if anything is happening there.
How many connections are there between graylog and mongo.
Has there been a configuration chance that could cause things running wild?
Is network okay between the cluster nodes?
Is mongod automatically updated?
Is java updated, i have seen things coming by with java updates causing strange things.
Is your ES cluster green?

Kind greetings,
Arie