Hey Logik folks,
I’m setting up Flame on AWS and have got my ec2 instance up and running with the converted Autodesk-supplied Rocky Linux 8.7 AMI, installed DCV and can connect, all good up until that point.
For storage, I’m using FSx for Lustre because that’s what I’m already using with Ubuntu on AWS. It should be fast enough.
But when I try to install the FSx for Lustre client on the Rocky Linux instance, it dies on reboot. Can no longer ssh in. If I don’t reboot after installing, but try mounting the filesystem with systemctl, I get a soft lockup with this spitting out to the shell:
Message from syslogd@flame01 at Jul 17 07:15:46 ... kernel:watchdog: BUG: soft lockup - CPU#3 stuck for 22s! [pmdalinux:27713]
And a reboot kills the instance and can no longer connect to it.
I’m following the instructions here:
…which worked previously just fine when I installed the client on an Ubuntu instance. That doc says my Rocky kernel 8.7 should be compatible. Hmm.
Anyone been through this already? I’ll post on the AWS forum too but thought I’d check here first.
Thanks
Ean