Elleni Veteran
Joined: 23 May 2006 Posts: 1289
|
Posted: Wed Aug 14, 2024 6:51 am Post subject: [solved] redis OOM crash while running gvm under high load |
|
|
I dedicated 100 GB RAM to this VM, and 40 vCPU's as I knew that gvm will need many ressources. Then while testing I did some simoultanious scans - like 4 to 5 x /24 networkscans. It mostly worked while the box was experiencing heavy load.
However - I saw some scans not finishing but being interrupted at 98%.
Looking through the logs, it became clear, that the redis-openvas service crashed because of oom. A restart of the service redis-openvas restored scanning functionality.
dmesg -T shows:
Code: |
[Di, 13. Aug 2024, 23:06:06] oom-kill:constraint=CONSTRAINT_NONE,nodemask=(null),cpuset=openrc.redis-openvas,mems_allowed=0,global_oom,task_memcg=/openrc.redis-openvas,task=redis-server,pid=2611,uid=495
[Di, 13. Aug 2024, 23:06:06] Out of memory: Killed process 2611 (redis-server) total-vm:155556648kB, anon-rss:99929064kB, file-rss:4896kB, shmem-rss:0kB, UID:495 pgtables:257052kB oom_score_adj:0
|
While I will for sure better plan scans in order to not overload the vm, but is there also a technical way to fix this "issue"? Can redis be limited somehow in its memory consumption?
Last edited by Elleni on Tue Aug 20, 2024 9:47 am; edited 1 time in total |
|