On Wed, Sep 25, 2024 at 08:56:57PM +0200, Filip Stojić wrote: > Hi, > > Don't know where or who to contact anymore. Found this email on > https://xfs.wiki.kernel.org/ > > I got an almalinux server with /var on a soft raid5 6x8tb disks. So a custom distro kernel - we're not really going to be able to help you much beyond basic triage. You'll need to talk to your distro vendor support channels for anything more than that. > There was a power failure and now on boot it states this: > > XFS (md3): Starting recovery (logdev: internal) > > Eventually it times out and continues the rest of the boot which kicks it > into emergency mode. (I guess because it doesn't have /var mounted) and > mount continues running in the background using 100% of 1 core. > > This has now been running for 215+ hours. Is this ok? That's not expected behaviour. Maybe tens of minutes given the slow storage and the possibility of thousands of inodes to recover, but I wouldn't expect /var to have that much to recover in it. > Anyone having any information on this would be greatly appreciated. several repeated sysrq-l dmesg output dumps from emergnecy mode will give us stack traces that tell us what recovery is doing. That might give us some clues as to what is going on. -Dave. -- Dave Chinner david@xxxxxxxxxxxxx
