On 2018年08月03日 12:55, Nathan Dehnel wrote: > Question 1: The command has been running for a couple days on a 10TB > array with no visible change in output. Is it actually fixing anything? > > enabling repair mode > Checking filesystem on > /dev/disk/by-uuid/e1ee5980-c54b-4b6e-82e2-3dbdcee1dd24 > UUID: e1ee5980-c54b-4b6e-82e2-3dbdcee1dd24 > checking extents > Fixed 0 roots. > checking free space cache > cache and super generation don't match, space cache will be invalidated > checking fs roots > root 5 inode 5043370 errors 100, file extent discount > Found file extent holes: > start: 0, len: 50331648 If none of above numbers changes, it's a dead loop. [snip] > > Question 2: I accidentally interrupted the command by logging out. How > dangerous is this? Should I be worried? Not more dangerous than --repair itself. As normally we do repair following metadata CoW, so it should be OK to interrupt. You should worry the first time you run --repair, other than worrying about interruption. Thanks, Qu
Attachment:
signature.asc
Description: OpenPGP digital signature
