Hi Otto,
Thanks for your reply.
On Thu, Jan 14, 2021 at 08:22:33AM +0100, Otto Moerbeek wrote:
| > Could there be some TOCTOU issue here somewhere? Or some cache
| > misbehaviour? Or is it really dying hardware?
|
| My first bet would be some form of corruption. FLipped bits in e..g
| directories while operating normally cannot be seen by the
| clean/unclean flag in the superblock. That one only records if the
| filesystem was unmounted before reboot, shutdown or crash.
I understand that - but then why would the error clear on subsequent
runs of rm?
| The forced fsck might reveal more.
It did find some issues, and then was waiting for my input over night
(when the backup run mounted the filesystem and changed things).
** /dev/sd2a (ebb54a869d056df3.a)
** File system is already clean
** Last Mounted on /backup
** Phase 1 - Check Blocks and Sizes
** Phase 2 - Check Pathnames
** Phase 3 - Check Connectivity
** Phase 4 - Check Reference Counts
ZERO LENGTH DIR I=57604332 OWNER=root MODE=40755
SIZE=0 MTIME=Jan 13 13:56 2021
CLEAR? [Fyn?] y
** Phase 5 - Check Cyl groups
FREE BLK COUNT(S) WRONG IN SUPERBLK
SALVAGE? [Fyn?] y
SUMMARY INFORMATION BAD
SALVAGE? [Fyn?] y
BLK(S) MISSING IN BIT MAPS
SALVAGE? [Fyn?] y
27766624 files, 396630326 used, 267754002 free (2016066 frags,
33217242 blocks, 0.3% fragmentation)
***** FILE SYSTEM WAS MODIFIED *****
I ran it once more after that, more issues were found:
** /dev/sd2a (ebb54a869d056df3.a)
** File system is already clean
** Last Mounted on /backup
** Phase 1 - Check Blocks and Sizes
** Phase 2 - Check Pathnames
** Phase 3 - Check Connectivity
** Phase 4 - Check Reference Counts
** Phase 5 - Check Cyl groups
FREE BLK COUNT(S) WRONG IN SUPERBLK
SALVAGE? [Fyn?] y
SUMMARY INFORMATION BAD
SALVAGE? [Fyn?] y
BLK(S) MISSING IN BIT MAPS
SALVAGE? [Fyn?] y
27884252 files, 397169471 used, 267214857 free (1944825 frags,
33158754 blocks, 0.3% fragmentation)
***** FILE SYSTEM WAS MODIFIED *****
Until the third fsck came back clean:
** /dev/sd2a (ebb54a869d056df3.a)
** File system is already clean
** Last Mounted on /backup
** Phase 1 - Check Blocks and Sizes
** Phase 2 - Check Pathnames
** Phase 3 - Check Connectivity
** Phase 4 - Check Reference Counts
** Phase 5 - Check Cyl groups
27884252 files, 397169471 used, 267214857 free (1944825 frags,
33158754 blocks, 0.3% fragmentation)
136m19.01s real 4m00.56s user 20m33.85s system
I'll write it off to those errors, but I still don't understand why
re-trying would fix these kinds of issues.
Thanks again, Otto!
Paul
--
>++++++++[<++++++++++>-]<+++++++.>+++[<------>-]<.>+++[<+
+++++++++++>-]<.>++[<------------>-]<+.--------------.[-]
http://www.weirdnet.nl/
