| LinHES Forums http://forums.linhes.org/ |
|
| XFS maintenance - read if you're new to XFS http://forums.linhes.org/viewtopic.php?f=3&t=18601 |
Page 1 of 1 |
| Author: | graysky [ Sun Jul 20, 2008 12:33 pm ] |
| Post subject: | XFS maintenance - read if you're new to XFS |
First off, here are two related threads for those of us new to XFS: How and why - create an XFS partition How to defragment XFS partitions and keep them that way You must periodically check your XFS partition(s) for error manually since fsck will NOT run automatically as it does with your ext3 partition(s). There are two utils on your system to do this: xfs_check and xfs_repair Checking for Inconsistencies Note that you MUST unmount your XFS partition to use either of these. Do NOT run either of them on a mounted filesystem! In my system, only my /myth is XFS (it is /dev/hda3). In order to unmount your partition, you must exit myth-frontend since it'll be using it. After you exited the frontend and/or mythwelcome, drop into single user mode and umount your partition (again mine is /dev/hda3 but substitute your XFS partition for hda3): Code: # telinit 1 # umount /dev/hda3 If you get a "umount: /myth: device is busy" error, do this: Code: $ fuser -m /dev/hda3 /dev/hda3: 3560 Mine returned one process (3560) using it. You can see what it is by issueing the following: Code: $ ps auxw|grep 3560 mythtv 3560 4.5 11.1 164128 53660 ? SLl 14:17 0:03 mythfrontend mythtv 3594 0.0 0.1 1780 596 ttyp0 S+ 14:19 0:00 grep 3560 It's mythfrontend (I didn't exit to illustrate this point). I killed it using the kill command followed-up a umount: Code: # kill 3560 # umount /dev/hda3 Now you can run xfs_check or xfs_repair safely Code: # xfs_check /dev/hda3 # xfs_check /dev/hda3 If there is nothing wrong with the file system, you won't see anything. From the xfs_check man page" Quote: Any output that is produced when xfs_check is not run in verbose mode indicates that the filesystem has an inconsistency. If you get no errors, just mount the partition and return to multiuser mode. Code: # mount /dev/hda3 # telinit 5 $ pkill xinit Attempt to Repair Inconsistencies Again, ONLY do the following on an unmounted partition! If you do get errors, you can run xfs_repair: Code: # xfs_repair /dev/hda3 Phase 1 - find and verify superblock... Phase 2 - using internal log - zero log... - scan filesystem freespace and inode maps... - found root inode chunk Phase 3 - for each AG... - scan and clear agi unlinked lists... - process known inodes and perform inode discovery... - agno = 0 - agno = 1 - agno = 2 - agno = 3 - process newly discovered inodes... Phase 4 - check for duplicate blocks... - setting up duplicate extent list... - check for inodes claiming duplicate blocks... - agno = 0 - agno = 1 - agno = 2 - agno = 3 Phase 5 - rebuild AG headers and trees... - reset superblock... Phase 6 - check inode connectivity... - resetting contents of realtime bitmap and summary inodes - traversing filesystem ... - traversal finished ... - moving disconnected inodes to lost+found ... Phase 7 - verify and correct link counts... done In my case, I had no errors, but wanted to show you the output of xfs_repair. So all is well. Now go back and remount your /dev/hda3 and swtich back to multiuser mode and finally restart your frontend with a pkill xinit: Code: # mount /dev/hda3
# telinit 5 $ pkill xinit Please see the man page for xfs_repair for a more detailed description of those Phases (referenced below). References docs.cray.com - Checking XFS File System Consistency with xfs_check and xfs_repair linux.die.net - man page for xfs_check and man page for xfs_repair ocaoimh.ie - How to umount when the device is busy |
|
| Author: | tjc [ Sun Jul 20, 2008 4:20 pm ] |
| Post subject: | Re: XFS maintenance - read if you're new to XFS |
graysky wrote: You must periodically check your XFS partition(s) for error manually since fsck will NOT run automatically as it does with your ext3 partition(s).
Well that's the best reason I've heard yet not to use XFS. Any FS developer who can't be bothered to make their FS utilities plug into the _very_ general and standard fsck framework just doesn't know how to play well with others. WTH were they thinking? 
|
|
| Author: | graysky [ Sun Jul 20, 2008 4:43 pm ] |
| Post subject: | |
I could be mistaken, but I read this in several places including within fsck when I attempted to use it on that partition: Code: fsck /dev/hda3
fsck 1.40.6 (09-Feb-2008) If you wish to check the consistency of an XFS filesystem or repair a damaged filesystem, see xfs_check(8) and xfs_repair(8). I was a bit surprised when I read this myself. But since I already made the entire /myth XFS I don't really care at this point. That said, from what I've read, either JFS or XFS are the recommended FS's to use when dealing with large files. |
|
| Author: | tjc [ Sun Jul 20, 2008 5:13 pm ] |
| Post subject: | |
I'd already done that investigation myself and verified that it was indeed the case. My response was as much an exclamation of disgust and annoyance with developers who can't be bothered to follow standards as anything. This is like a stereo receiver which requires some bizarre custom connectors instead any of the standard types (RCA phone, bare wire, banana plugs, ...) You'd take one look and say "WTH were they thinking?!?" and spend your money elsewhere if you weren't a total tech victim. http://ars.userfriendly.org/cartoons/?id=20080618 |
|
| Author: | graysky [ Sun Jul 20, 2008 5:34 pm ] |
| Post subject: | |
@tjc - too right! |
|
| Page 1 of 1 | All times are UTC - 6 hours |
| Powered by phpBB® Forum Software © phpBB Group http://www.phpbb.com/ |
|