On Wed, 11 Nov 2020 17:48:33 +0100 "Aleksej Lebedev" <[email protected]> wrote:
> ... > The only thing is that it requires copying >0.5TB. I am still not sure > it's 100% reproducible, but for me I got these errors 3 times in a row. > At different stages though. Hm, this sounds similar to what I've been bitten several times when running periodic and buildworld simultaneously, see my post from September: http://lists.dragonflybsd.org/pipermail/users/2020-September/381405.html I was not sure whether to blame bulkfree or snapshot, but it well may be the same issue you are experiencing, ie. bulkfree combined with massive filesystem activity. And impact was the same for me - several "check fail" errors from kernel, and couple of undeletable entries on the affected filesystem. If it really is the same issue, you'd not need to copy that amount of data, just run "make build-all" (ie. parallelized build) on hammer2 filesystem and bulkfree at the same time. -- Daniel
