Over the course of the last few months on a good 8-10 of differing application servers we have which run CentOS 6.10, 2.6.32-754 as OS and Kernel revision we are getting into a
KERNEL PANIC – not syncing: Out of Memory and no killable processes event
if we catch a slowing VM prior to the Kernel Panic The /proc/meminfo output shows almost all the RAM is stuck in "SUnreclaim"
e.g.
cat /proc/meminfo | grep "Slab\|SReclaim\|SUnreclaim"
Slab: 1902332 kB
SReclaimable: 12500 kB
SUnreclaim: 1889832 kB
Is there a way to find out which processes are leaking their ram away into SUnreclaim so i can chase / hunt down the offending application or process tree?
Thanks in anticipation for any help that can be provided.
AK@Work