I have a couple of RAID arrays attached to an old(-ish) LSI 9285CV-8e
controller and, while doing a backup between them, I seem to be running
into a driver/controller resource issue :-(
Every 5 minutes when a 'health check' runs, it logs a shower of "mrsas0:
Cannot allocate ioctl data mem" messages. This seems to be as a result
of running into the 16 s/g buffer limit for pass-through IOCTLs used by
smartd and/or storcli.
There doesn't appear to be any data corruption, just annoying log lines.
Has anyone else run into this?
Michael