ovh / beamium

Prometheus to Warp10 metrics forwarder
Other
84 stars 20 forks source link

100% CPU on CentOS 7 using most of the RAM loadavg 90 #137

Open wwalker opened 3 years ago

wwalker commented 3 years ago

Load average hit 90 before I got a successful ssh session and find out what was killing the machine.

It looks like it started leaking memory 56 hours before the load average went through the roof.

In that time it leaked about 6.5 GB (on a machine with only 8 GB, and essentially idle) Screenshot_2021-10-27_22-10-23

The load went from 0.1 to 90.0 in 3 hours and 15 minutes. Screenshot_2021-10-27_22-14-28

All the problems went away after pkill -f beamium