Re: [Ocfs2-users] ocfs2 problem journal size
The 4 journal inodes got zeroed out. Do you know how/why? Have you tried running fsck with -fy (enable writes). fsck.ocfs2 does have a check for bad journals that it will regenerate. JOURNAL_FILE_INVALID OCFS2 uses JDB for journalling and some journal files exist in the system directory. Fsck has found some journal files that are invalid. Answering yes to this question will regenerate the invalid journal files. But that may still not work as fsck is currently bailing out during journal recovery that happens much earlier on. Try with -fy. If that does not work, we'll have to reconstruct empty inodes as placeholders to allow fsck to complete journal recovery followed by journal recreation. On Wed, Aug 1, 2012 at 6:41 PM, Christophe BOUDER christophe.bou...@lip6.fr wrote: Hello, i use ocfs2 1.6.3 kernel 3.4.4 on debian testing i had problem on my infortrend device media error on a disk the result i can't mount my ocfs2 file but i can read the files with debugfs.ocfs2 and my question is can i recover or recreate the journal size for node 8 9 10 11 ? thank for your help here's some log : # mount /data mount.ocfs2: Internal logic failure while trying to join the group # fsck.ocfs2 -n /dev/sdc1 fsck.ocfs2 1.6.3 Checking OCFS2 filesystem in /dev/sdc1: Label: data UUID: 9B655B51E6874480BBC1309DCA048A39 Number of blocks: 4027690992 Block size: 4096 Number of clusters: 251730687 Cluster size: 65536 Number of slots:32 journal recovery: I/O error on channel while reading cached inode 112 for slot 8's journal fsck encountered unrecoverable errors while replaying the journals and will not continue # echo ls -l // | debugfs.ocfs2 /dev/sdc1 |grep journal debugfs.ocfs2 1.6.3 55 -rw-r--r-- 1 0 0 268435456 23-Jun-2007 21:30 journal: 56 -rw-r--r-- 1 0 0 268435456 23-Jun-2007 21:30 journal:0001 57 -rw-r--r-- 1 0 0 268435456 23-Jun-2007 21:30 journal:0002 58 -rw-r--r-- 1 0 0 268435456 23-Jun-2007 21:30 journal:0003 59 -rw-r--r-- 1 0 0 268435456 23-Jun-2007 21:31 journal:0004 79 -rw-r--r-- 1 0 0 268435456 31-Aug-2007 00:45 journal:0005 80 -rw-r--r-- 1 0 0 268435456 31-Aug-2007 00:45 journal:0006 81 -rw-r--r-- 1 0 0 268435456 31-Aug-2007 00:45 journal:0007 112 -- 0 0 0 0 1-Jan-1970 01:00 journal:0008 113 -- 0 0 0 0 1-Jan-1970 01:00 journal:0009 114 -- 0 0 0 0 1-Jan-1970 01:00 journal:0010 115 -- 0 0 0 0 1-Jan-1970 01:00 journal:0011 116 -rw-r--r-- 1 0 0 268435456 31-Aug-2007 00:46 journal:0012 117 -rw-r--r-- 1 0 0 268435456 31-Aug-2007 00:47 journal:0013 118 -rw-r--r-- 1 0 0 268435456 31-Aug-2007 00:47 journal:0014 119 -rw-r--r-- 1 0 0 268435456 31-Aug-2007 00:47 journal:0015 142 -rw-r--r-- 1 0 0 268435456 29-May-2009 22:53 journal:0016 143 -rw-r--r-- 1 0 0 268435456 29-May-2009 22:54 journal:0017 166 -rw-r--r-- 1 0 0 268435456 31-Jan-2010 15:36 journal:0018 167 -rw-r--r-- 1 0 0 268435456 31-Jan-2010 15:36 journal:0019 168 -rw-r--r-- 1 0 0 268435456 31-Jan-2010 15:37 journal:0020 169 -rw-r--r-- 1 0 0 268435456 31-Jan-2010 15:37 journal:0021 170 -rw-r--r-- 1 0 0 268435456 31-Jan-2010 15:38 journal:0022 171 -rw-r--r-- 1 0 0 268435456 31-Jan-2010 15:38 journal:0023 208 -rw-r--r-- 1 0 0 268435456 21-Nov-2010 19:35 journal:0024 209 -rw-r--r-- 1 0 0 268435456 21-Nov-2010 19:35 journal:0025 210 -rw-r--r-- 1 0 0 268435456 21-Nov-2010 19:36 journal:0026 211 -rw-r--r-- 1 0 0 268435456 21-Nov-2010 19:36 journal:0027 212 -rw-r--r-- 1 0 0 268435456 21-Nov-2010 19:36 journal:0028 213 -rw-r--r-- 1 0 0 268435456 21-Nov-2010 19:36 journal:0029 214 -rw-r--r-- 1 0 0 268435456 21-Nov-2010 19:37 journal:0030 215 -rw-r--r-- 1 0 0 268435456
Re: [Ocfs2-users] ocfs2 problem journal size
hello, The 4 journal inodes got zeroed out. Do you know how/why? raid6 with 2 bad disk and a third who got problem reinsert it in the device it appears good but it also crash the device not recognize by the system. Have you tried running fsck with -fy (enable writes). yes but without success #fsck.ocfs2 -fy /dev/sdc1 fsck.ocfs2 1.6.3 fsck.ocfs2: Internal logic failure while initializing the DLM Try with -fy. If that does not work, we'll have to reconstruct empty inodes as placeholders to allow fsck to complete journal recovery followed by journal recreation. ok how can i do that ? -- Christophe Bouder, ___ Ocfs2-users mailing list Ocfs2-users@oss.oracle.com https://oss.oracle.com/mailman/listinfo/ocfs2-users
Re: [Ocfs2-users] ocfs2 problem journal size
oh crap. The dlm lock needs to lock the journals. So you need to recreate the journal inodes with i_size 0. dd a good journal inode and edit it using binary editor. Change the inode num to the block number, zero out the i_size and next_free_extent. Repeat for the 4 inodes. Hopefully some one on the list has the time to help you further. On Thu, Aug 2, 2012 at 10:50 AM, Christophe BOUDER christophe.bou...@lip6.fr wrote: hello, The 4 journal inodes got zeroed out. Do you know how/why? raid6 with 2 bad disk and a third who got problem reinsert it in the device it appears good but it also crash the device not recognize by the system. Have you tried running fsck with -fy (enable writes). yes but without success #fsck.ocfs2 -fy /dev/sdc1 fsck.ocfs2 1.6.3 fsck.ocfs2: Internal logic failure while initializing the DLM Try with -fy. If that does not work, we'll have to reconstruct empty inodes as placeholders to allow fsck to complete journal recovery followed by journal recreation. ok how can i do that ? -- Christophe Bouder, ___ Ocfs2-users mailing list Ocfs2-users@oss.oracle.com https://oss.oracle.com/mailman/listinfo/ocfs2-users
