Hi
a box i have crashed this morning, the problem is that :
# raidstart -a
# cat /proc/kmsg
<4>(read) hdb1's sb offset: 15016576 [events: 0000004a]
<4>(read) hde1's sb offset: 15016576 [events: 0000004a]
<4>(read) hdf1's sb offset: 15016576 [events: 0000004a]
<4>(read) hdg1's sb offset: 15016576 [events: 00000048]
<4>(read) hdh1's sb offset: 15016576 [events: 00000049]
<4>autorun ...
<4>considering hdh1 ...
<4> adding hdh1 ...
<4> adding hdg1 ...
<4> adding hdf1 ...
<4> adding hde1 ...
<4> adding hdb1 ...
<4>created md0
<4>bind<hdb1,1>
<4>bind<hde1,2>
<4>bind<hdf1,3>
<4>bind<hdg1,4>
<4>bind<hdh1,5>
<4>running: <hdh1><hdg1><hdf1><hde1><hdb1>
<4>now!
<4>hdh1's event counter: 00000049
<4>hdg1's event counter: 00000048
<4>hdf1's event counter: 0000004a
<4>hde1's event counter: 0000004a
<4>hdb1's event counter: 0000004a
<3>md: superblock update time inconsistency -- using the most recent one
<4>freshest: hdf1
<4>md: kicking non-fresh hdg1 from array!
<4>unbind<hdg1,4>
<4>export_rdev(hdg1)
<4>md0: removing former faulty hdg1!
<4>md0: kicking faulty hdh1!
<4>unbind<hdh1,3>
<4>export_rdev(hdh1)
<3>md: md0: raid array is not clean -- starting background reconstruction
<6>raid5 personality registered
<6>md0: max total readahead window set to 512k
<6>md0: 4 data-disks, max readahead per data-disk: 128k
<6>raid5: device hdf1 operational as raid disk 2
<6>raid5: device hde1 operational as raid disk 1
<6>raid5: device hdb1 operational as raid disk 0
<3>raid5: not enough operational devices for md0 (2/5 failed)
<4>RAID5 conf printout:
<4> --- rd:5 wd:3 fd:2
<4> disk 0, s:0, o:1, n:0 rd:0 us:1 dev:hdb1
<4> disk 1, s:0, o:1, n:1 rd:1 us:1 dev:hde1
<4> disk 2, s:0, o:1, n:2 rd:2 us:1 dev:hdf1
<4> disk 3, s:0, o:0, n:3 rd:3 us:1 dev:[dev 00:00]
<4> disk 4, s:0, o:0, n:4 rd:4 us:1 dev:[dev 00:00]
<4> disk 5, s:0, o:0, n:0 rd:0 us:0 dev:[dev 00:00]
<4> disk 6, s:0, o:0, n:0 rd:0 us:0 dev:[dev 00:00]
<4> disk 7, s:0, o:0, n:0 rd:0 us:0 dev:[dev 00:00]
<4> disk 8, s:0, o:0, n:0 rd:0 us:0 dev:[dev 00:00]
<4> disk 9, s:0, o:0, n:0 rd:0 us:0 dev:[dev 00:00]
<4> disk 10, s:0, o:0, n:0 rd:0 us:0 dev:[dev 00:00]
<4> disk 11, s:0, o:0, n:0 rd:0 us:0 dev:[dev 00:00]
<1>raid5: failed to run raid set md0
<4>pers->run() failed ...
<4>do_md_run() returned -22
<4>unbind<hdf1,2>
<4>export_rdev(hdf1)
<4>unbind<hde1,1>
<4>export_rdev(hde1)
<4>unbind<hdb1,0>
<4>export_rdev(hdb1)
<6>md0 stopped.
<4>... autorun DONE.
how could i tell the kernel to start the raid array and do what it can to
recover the datas ?
as far as i can tell, the problem comes from :
<4>hdh1's event counter: 00000049
<4>hdg1's event counter: 00000048
<4>hdf1's event counter: 0000004a
<4>hde1's event counter: 0000004a
<4>hdb1's event counter: 0000004a
how could i modify the event counter of one of the disks to get it work ?
--
Mathieu Arnold
