We’re trying to deploy Rocky Linux on a cluster of old hardware that requires aacraid for the disk controller. Installing v8.9 with the dd-aacraid DUD image from ELRepo gives us a system that appears to function properly up until the point of attempting a reboot (of the installed system), at which instance we keep getting ‘aacraid: Outstanding commands on …’, making the system no longer usable.
Researching, and also comparing the Rocky Linux 8 system to a Debian 12 one that does not seem to suffer from that issue, we configured a udev rule to set /dev/block/sda/device/timeout to a value of 60 rather than the Rocky Linux default of 45, but that still did not help.
I was wondering if there could be additional suggestions we could try.
P.S. We’re still installing using the v8.9 OS image, because v8.10 with the corresponding dd-aacraid from ELRepo does not even detect the drive, details here: