Description

After running a full fsck to repair a corrupted filesystem,fsck fails with no valid OLT

# fsck -F vxfs -o full  /dev/vx/rdsk/testdg/testvol  
vxfs fsck: ERROR: V-3-20726: OLT extent 0 has bad checksum read of primary OLT failed
UX:vxfs fsck: ERROR: V-3-20726: OLT extent 1 has bad checksum read of OLT copy failed
UX:vxfs fsck: ERROR: V-3-20718: no valid OLT, cannot continue file system check failure, aborting

1.  check the volume size of the effect filesystem

# vxprint –g testdg testvol

v   tvol          -             ENABLED   ACTIVE    17616076 SELECT     -        fsgen
pl tvol-01       tvol          ENABLED   ACTIVE    17617920 CONCAT     -        RW
sd d1-01         tvol-01       d1        0         17617920 0          c4t5d30 ENA

Volume size = 17616076 sectors

2. Validate the superblock using the fsdb filesystem debugging tool

# fsdb   -F vxfs /dev/vx/rdsk/testdg/testvol

  > 8192B.p S

super-block at 00000008.0000
magic a501fcf5  version 6
ctime 1096293624 999550  ( Mon Sep 27 07:00:24 2004 PDT )      << check create time
log_version 12 logstart 0  logend 0
bsize  1024 size  8807078 dsize  8807078  ninode 0  nau 0     << use bsize & size values to match volume size *
defiextsize 0  oilbsize 0  immedlen 96  ndaddr 10
aufirst 0  emap 0  imap 0  iextop 0  istart 0
bstart 0  femap 0  fimap 0  fiextop 0  fistart 0  fbstart 0
nindir 2048  aulen 32768  auimlen 0  auemlen 8
auilen 0  aupad 0  aublocks 32768  maxtier 15
inopb 4  inopau 0  ndiripau 0  iaddrlen 8   bshift 10
inoshift 2  bmask fffffc00  boffmask 3ff  checksum e66ac5be
free 3326029  ifree 0
efree  2157 2422 3099 729 426 229 198 148 97 73 30 12 10 8 5 18 12 2 3 1 0 0 0 0 0 0 0 0
flags 0 mod 1 clean 3c
time 1314698781 367816  (Tue Aug 30 03:06:21 2011 PDT)    << check last superblock flush is current
oltext[0] 45958  oltext[1] 46019  oltsize 1                           << location of the 2 OLT copies if they exist
iauimlen 1  iausize 4  dinosize 256                      
checksum2 0  checksum3 0
fsetquotaction 0
fsetquotahardlimit 0 fsetquotasoftlimit 0
log_gen 15

 *  checking the sizes of the filesystem and volume match

where

filesystem size in sectors = ( bsize*size)/512 = 1024*8807078 = 17616076       << size match the volume size

> olt

failed reading the OLT             <<  as expected since there are no valid olt

2.   display the content of the OLT blocks at   both locations as specified in the superblock above

>   45958b.p 100 c

0.0000b386.0000:  t  h  i  s     i  s      a  d  a  t  a      f
0.0000b386.0010:  r  o  m      a      u  s  e  r         a  n  d
0.0000b386.0020:  n  o  t         m  e  t  a  d  a  t  a      s   o
0.0000b386.0030:  _  i  t     a  p  p  e  a  r  s      d  a  t  a
0.0000b386.0040:  h  a  s  b  e  e  n      o  v  e   r      w  r  i
0.0000b386.0050:  t   t  e  n      t   h  e      m   e   t   a  d   a  t
0.0000b386.0060:  a   . . .   . .

> 46019b.p 100 c

0.0000b386.0000:   m   o   r   e     u   s  e   r   d  a  t  a      a   n
0.0000b386.0010:  d  a  g  a  i  n   u  s  e  r      d  a  t  a
0.0000b386.0020:  n  o  t         m  e  t  a  d  a  t  a      s   o
0.0000b386.0030:  _  i  t     a  p  p  e  a  r  s      d  a  t  a
0.0000b386.0040:  h  a  s  b  e  e  n      o  v  e   r      w  r  i
0.0000b386.0050:  t   t  e  n      t   h  e      m   e   t   a  d   a  t
0.0000b386.0060:  a   . . .   . .

> quit

It is clear this does not contain OLT data since it has been overwritten by user data

For a valid OLT we would see an output similar to this

> 45958b.p 100 c
0.00000020.0000: a5 04 fc f5 00 00 00 @ N i be 8d 00 00 00 01          << OLT magic number=a504fcf5 exist at start
0.00000020.0010: N e 91 10 00 04 13 8 00 00 00 00 00 00 00 00
0.00000020.0020: 00 00 00 00 00 00 09 02 00 00 00 00 00 00 00 00
0.00000020.0030: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 03 08
0.00000020.0040: 00 00 00 02 00 00 00 18 00 00 00 00 00 00 00 03
0.00000020.0050: 00 00 00 00 00 00 00 # 00 00 00 04 00 00 00 18
0.00000020.0060: 00 00 00 00
 

3.   Check if the OLT may be located at a different location using the OLT magic no a504fcf5

#   dd if=/dev/vx/rdsk/testdg/tvol |od -Ax -X | grep a504fcf5

If it fails to return it indicates there is no copy of the OLT and there is no chance of recovery of this filesystem.   

But if the dd commands returns a valid OLT then follow the process as per the article: https://support-staff.cloud.com/wolken/esd/knowledge-base-view/view-kb-article?articleNumber=100020562&isLocationBackOnHome=true&hideTabs=true 

 

Note: This is valid for all platforms and in this example, we have used Solaris, for other platforms, replace "-F vxfs" as follows:   

               Linux : -t vxfs  ;           AIX: -V vxfs ;       HPUX: -F vxfs

How to check if a valid OLT exist on a VXFS filesystem