Problem restarting gvinum raid-5
glz
goran.lowkrantz at ismobile.com
Fri Jul 7 14:35:55 UTC 2006
First, missed some info:
> uname -a
FreeBSD byleist.hq.ismobile.com 6.1-STABLE FreeBSD 6.1-STABLE #1: Mon
Jun 26 20:37:45 CEST 2006
root at byleist.hq.ismobile.com:/usr/obj/usr/src/sys/BYLEISTSMP i386
So I continued to dig into this and it seems that the plex is not
rebuilding because the geom is open, i.e. this routine:
/* Check if any consumer of the given geom is open. */
int
gv_is_open(struct g_geom *gp)
{
struct g_consumer *cp;
if (gp == NULL)
return (0);
LIST_FOREACH(cp, &gp->consumer, consumer) {
if (cp->acr || cp->acw || cp->ace)
return (1);
}
return (0);
}
What does this mean? How do I make sure the geom is not opened until I
can start the plex? I have tested single user and not mounted file
system but it does not help.
From the code I have read, this state should be the proper for allowing
a rebuild of the plex:
5 drives:
D disk5 State: up /dev/da6s1a A: 0/17492 MB (0%)
D disk4 State: up /dev/da5s1a A: 0/17492 MB (0%)
D disk3 State: up /dev/da4s1a A: 0/17492 MB (0%)
D disk2 State: up /dev/da3s1a A: 0/17492 MB (0%)
D disk1 State: up /dev/da2s1a A: 0/17492 MB (0%)
1 volume:
V imap State: up Plexes: 1 Size: 68 GB
1 plex:
P imap.p0 R5 State: degraded Subdisks: 5 Size: 68 GB
5 subdisks:
S imap.p0.s0 State: up D: disk1 Size: 17 GB
S imap.p0.s1 State: up D: disk2 Size: 17 GB
S imap.p0.s2 State: up D: disk3 Size: 17 GB
S imap.p0.s3 State: up D: disk4 Size: 17 GB
S imap.p0.s4 State: stale D: disk5 Size: 17 GB
/glz
Goran Lowkrantz wrote:
> Hi,
>
> We have a gvinum raid-5 volume that that we had to replace a disk on and
> after that we cant get the new subdisk starting.
>
> Here are the things we did:
> 1: Replace disk and boot singleuser to fdisk and lable new disk:
> gvinum -> list
> 5 drives:
> D disk4 State: up /dev/da5s1a A: 0/17492 MB (0%)
> D disk3 State: up /dev/da4s1a A: 0/17492 MB (0%)
> D disk2 State: up /dev/da3s1a A: 0/17492 MB (0%)
> D disk1 State: up /dev/da2s1a A: 0/17492 MB (0%)
>
> 1 volume:
> V imap State: up Plexes: 1 Size: 68 GB
>
> 1 plex:
> P imap.p0 R5 State: up Subdisks: 5 Size: 68 GB
>
> 5 subdisks:
> S imap.p0.s0 State: up D: disk1 Size: 17 GB
> S imap.p0.s1 State: up D: disk2 Size: 17 GB
> S imap.p0.s2 State: up D: disk3 Size: 17 GB
> S imap.p0.s3 State: up D: disk4 Size: 17 GB
> S imap.p0.s4 State: up D: disk5 Size: 17 GB
>
> After fixing the new disk partition we did a saveconfig and reboot:
> gvinum -> list
> 5 drives:
> D disk5 State: up /dev/da6s1a A: 0/17492 MB (0%)
> D disk4 State: up /dev/da5s1a A: 0/17492 MB (0%)
> D disk3 State: up /dev/da4s1a A: 0/17492 MB (0%)
> D disk2 State: up /dev/da3s1a A: 0/17492 MB (0%)
> D disk1 State: up /dev/da2s1a A: 0/17492 MB (0%)
>
> 1 volume:
> V imap State: up Plexes: 1 Size: 68 GB
>
> 1 plex:
> P imap.p0 R5 State: up Subdisks: 5 Size: 68 GB
>
> 5 subdisks:
> S imap.p0.s4 State: stale D: disk5 Size: 17 GB
> S imap.p0.s3 State: up D: disk4 Size: 17 GB
> S imap.p0.s2 State: up D: disk3 Size: 17 GB
> S imap.p0.s1 State: up D: disk2 Size: 17 GB
> S imap.p0.s0 State: up D: disk1 Size: 17 GB
>
> Tried start on plex and subdisk, nnot working. Finally, to get plex into
> degraded mode we did a setstate down imap.p0.s4.
> gvinum -> list
> 5 drives:
> D disk5 State: up /dev/da6s1a A: 0/17492 MB (0%)
> D disk4 State: up /dev/da5s1a A: 0/17492 MB (0%)
> D disk3 State: up /dev/da4s1a A: 0/17492 MB (0%)
> D disk2 State: up /dev/da3s1a A: 0/17492 MB (0%)
> D disk1 State: up /dev/da2s1a A: 0/17492 MB (0%)
>
> 1 volume:
> V imap State: up Plexes: 1 Size: 68 GB
>
> 1 plex:
> P imap.p0 R5 State: degraded Subdisks: 5 Size: 68 GB
>
> 5 subdisks:
> S imap.p0.s4 State: down D: disk5 Size: 17 GB
> S imap.p0.s3 State: up D: disk4 Size: 17 GB
> S imap.p0.s2 State: up D: disk3 Size: 17 GB
> S imap.p0.s1 State: up D: disk2 Size: 17 GB
> S imap.p0.s0 State: up D: disk1 Size: 17 GB
>
> and here we are. Start on volume or plex give errno 16, start on subdisk
> gives can't start: cannot start 'imap.p0.s4' - not yet supported.
>
> Can't find any descriptions of the proper way to do disk replacement, so
> if this is wrong, I'd love to get updated. And how do we get the current
> situation upa nd running?
>
> Regards,
> Göran
>
>
>
> ................................................... the future isMobile
>
> Goran Lowkrantz <goran.lowkrantz at ismobile.com>
> System Architect, isMobile, Aurorum 2, S-977 75 Luleå, Sweden
> Phone: +46(0)920-75559
> Mobile: +46(0)70-587 87 82 Fax: +46(0)70-615 87 82
>
> http://www.ismobile.com ...............................................
> _______________________________________________
> freebsd-stable at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-stable
> To unsubscribe, send any mail to "freebsd-stable-unsubscribe at freebsd.org"
--
................................................... the future isMobile
Goran Lowkrantz <goran.lowkrantz at ismobile.com>
System Architect, isMobile, Aurorum 2, S-977 75 Luleå, Sweden
Phone: +46(0)920-75559
Mobile: +46(0)70-587 87 82 Fax: +46(0)70-615 87 82
http://www.ismobile.com ...............................................
More information about the freebsd-stable
mailing list