Hi,

We have 2 nfs servers in our environment with config as mentioned below :

1)

OS : Rhel 7.4
Kernel : 3.10.0-693.1.1.el7.x86_64
Make : Dell Inc. PowerEdge R630
Storage: Dell array MD34xx

2)

OS : Rhel 8.2
Kernel : 4.18.0-193.el8.x86_64
Make : Dell Inc. PowerEdge R640
Storage: Dell array ME40xx

Both the servers had incident were the nfs file system formatted as xfs soared to 100%. I/O to both the servers were requested and were being servred by bpth of them when the file system was 100%. However the one running Rhel 8.2 crashed after some time due to kernel panic with xfs i/o error whereas the other Rhel 7.4 box kept running fine. XFS errors were gone once the server was rebooted, nfs service stopped, some space created and nfs service resumed. Crash logs are attached to this thread.

Can you'all share some insights by referring to attached logs on these different behavior ?

Attached Files

rhel_8_2_crash_logs.txt (17.3 KB, 22 views)

Hi,

Any inputs on what went wrong with the crashed server would help.

There is a bug in the kernel or the xfs driver.
Turn to the RedHat support!