Join GitHub today
GitHub is home to over 40 million developers working together to host and review code, manage projects, and build software together.Sign up
2247.3.0 "wrong slab cache" lockups, may be related to cephfs #2616
Container Linux Version
2247.3.0 beta channel
Bare metal, Dell PowerEdge r720
Machine should keep running longer than 5 minutes after booting 2247.3.0 with cephfs mounts
Machine starts logging many times per second:
With an occasional oops message. Eventually it reboots, possibly due to watchdog.
This may be related to cephfs mounts. I was able to downgrade to stable with this kernel after stopping the docker service and killing all cephfs mounts.
I'm sorry I don't have more information here. This machine needed to get back into service ASAP. Hopefully this bug report can be built up by other Container Linux users.
It just began today, it appears to be new with 2247.3.0 because when we reverted to the other partition there was no problem.
I'll make an attempt to reproduce this on a less-critical machine and get back to you. The machine with the problem should never have been on beta.
H I am experiencing the same issue here with CoreOS 2247.3.0 on a vmware host. The node crashes very often with reboots. It seems like it starts spamming Ceph with lots of requests. Reverting to stable and everything works fine.
Also having same issue with CoreOS 2247.3.0 running under KVM 2.12.0. Reverted to 2247.2.0 and issue does not occur.
Unlike @nealey though, I only got the