On Tue, Oct 09, 2018 at 06:00:51PM +0200, Steinar H. Gunderson wrote: > We had a power loss event, and when a server with dm-cache came up again, > it paniced (see below for the panic text). I couldn't find any other way to > remedy this than to blow away the metadata volume, which I assumed was safe > as the cache is in writethrough mode (after several catastrophic events with > dm-cache earlier, I don't trust writeback anymore). Unfortunately, this was > seemingly not enough, as the underlying devices came back with various levels > of corruption and eventually had to be restored from backup. (It's running > without dm-cache now.) Is there some other place I should report this? The machine is awfully slow without its cache, so it's sort of annoying to run without. :-) /* Steinar */ -- Homepage: https://www.sesse.net/ -- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
