[LUGOS] RAID

Nejc Skoberne nejc.skoberne at guest.arnes.si
Thu Jun 20 01:15:34 CEST 2002


Zdravo.

>> Kako torej rocno rekonstruiram md0? mkraid /dev/md0?
> ko bos pognal raidhotadd; ce ti bo pojamral disk busy, naredi to v single
> user nacinu (init 1).

Jun 20 01:08:32 Vangelis kernel: md: trying to hot-add sda1 to md0 ...
Jun 20 01:08:32 Vangelis kernel: md: bind<sda1,2>
Jun 20 01:08:34 Vangelis kernel: md: updating md0 RAID superblock on device
Jun 20 01:08:34 Vangelis kernel: md: sda1 [events: 0000001a]<6>(write) sda1's sb offset: 2008000
Jun 20 01:08:34 Vangelis kernel: md: sdb1 [events: 0000001a]<6>(write) sdb1's sb offset: 2008000
Jun 20 01:08:34 Vangelis kernel: md: recovery thread got woken up ...
Jun 20 01:08:34 Vangelis kernel: md0: resyncing spare disk sda1 to replace failed disk
Jun 20 01:08:37 Vangelis kernel: md: syncing RAID array md0
Jun 20 01:08:37 Vangelis kernel: md: minimum _guaranteed_ reconstruction speed: 100 KB/sec/disc.
Jun 20 01:08:37 Vangelis kernel: md: using maximum available idle IO bandwith (but not more than 100000 KB/sec) for reconstruction.
Jun 20 01:08:37 Vangelis kernel: md: using 124k window, over a total of 2008000 blocks.

root at Vangelis:~# cat /proc/mdstat
Personalities : [raid1]
read_ahead 1024 sectors
md0 : active raid1 sda1[2] sdb1[1]
      2008000 blocks [2/1] [_U]
      [=====>...............]  recovery = 28.2% (567872/2008000) finish=7.2min speed=3329K/sec
unused devices: <none>

Ocitno sedaj rekonstruira pravilno.

> pojavil kak bolj inteligenten algoritem na to temo. so tile scsi diski
> IBMovi? pod kaksnim loadom jih furas? so ustrezno hlajeni?

Western Digital diska sta. Pod nikakvim: load average: 1.02, 0.60, 0.24
Mislim, da niso cisto tako, kot bi morali, je pa res, da delajo pod
nizkimi obrati in da so malce starejsi. Sicer ne vem ce je to lajsalna
okoliscina. :)

-- 
Nejc Skoberne
Grajska ulica 5
SI-5220 Tolmin
E-mail: nejc.skoberne at guest.arnes.si




More information about the lugos-list mailing list