* Strange RAID2 behavier...
@ 2002-05-30 11:59 me
2002-05-30 12:37 ` me
0 siblings, 1 reply; 8+ messages in thread
From: me @ 2002-05-30 11:59 UTC (permalink / raw)
To: linux-kernel; +Cc: Andre Hedrick
Ive got strange raid2 behaiver happening, ive got this raidtab:
----- raidtab -----
raiddev /dev/md0
raid-level 0
nr-raid-disks 2
persistent-superblock 1
chunk-size 8
device /dev/hde1
raid-disk 0
device /dev/hdi1
raid-disk 1
raiddev /dev/md1
raid-level 0
nr-raid-disks 2
persistent-superblock 1
chunk-size 8
device /dev/hdf1
raid-disk 0
device /dev/hdj1
raid-disk 1
# not in use right now
#raiddev /dev/md2
# raid-level 0
# nr-raid-disks 2
# persistent-superblock 1
# chunk-size 8
# device /dev/hdg1
# raid-disk 0
# device /dev/hdk1
# raid-disk 1
raiddev /dev/md3
raid-level 0
nr-raid-disks 2
persistent-superblock 1
chunk-size 8
device /dev/hdh1
raid-disk 0
device /dev/hdl1
raid-disk 1
----- raidtab -----
Im missing one disk on the md2 so im not using that right now but that
should not affect the raid2 or make this happen:
# raidstop --all /dev/md3
/dev/md3: Device or resource busy
# cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid5] [multipath]
read_ahead 1024 sectors
md3 : active raid0 hdl1[1] hdh1[0]
240688000 blocks 8k chunks
unused devices: <none>
md0 and md1 gets turned off just fine, but not md3 ????
I will try and rebuild the raid and mode md3 disks to md2 and see what
happens.
# uname -a
Linux odd 2.4.19-pre9 #1 SMP Wed May 29 08:46:00 CEST 2002 i686 unknown
There is no new 2.4.19-pre patch out from what i can find.
Thanks in advance for the help.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: Strange RAID2 behavier...
2002-05-30 12:37 ` me
@ 2002-05-30 12:12 ` Lionel Bouton
2002-05-30 12:21 ` Lionel Bouton
2002-05-30 13:12 ` me
0 siblings, 2 replies; 8+ messages in thread
From: Lionel Bouton @ 2002-05-30 12:12 UTC (permalink / raw)
To: me; +Cc: linux-kernel, Andre Hedrick
On jeu, mai 30, 2002 at 02:37:53 +0200, me@vger.org wrote:
> I made the md2 a linear raid of one drive, now I can stop the md3.
> This means that for example making md0 and md3 will make md3 unstoppable,
> has this bug already been reported?
This could be a bug/constraint in raidtools as said in mad raidtab:
"the parsing code isn't overly bright".
I've had small glitches when using comments in raidtab. Try removing all
your commented /dev/md2 related lines instead of building a dumb array.
LB.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: Strange RAID2 behavier...
2002-05-30 12:12 ` Lionel Bouton
@ 2002-05-30 12:21 ` Lionel Bouton
2002-05-30 13:12 ` me
1 sibling, 0 replies; 8+ messages in thread
From: Lionel Bouton @ 2002-05-30 12:21 UTC (permalink / raw)
To: linux-kernel
On jeu, mai 30, 2002 at 02:12:25 +0200, Lionel Bouton wrote:
> On jeu, mai 30, 2002 at 02:37:53 +0200, me@vger.org wrote:
> > I made the md2 a linear raid of one drive, now I can stop the md3.
> > This means that for example making md0 and md3 will make md3 unstoppable,
> > has this bug already been reported?
>
> This could be a bug/constraint in raidtools as said in mad raidtab:
s/mad/man/
"mad raidtab" could have been an accurate description of the problem though
:-)
> "the parsing code isn't overly bright".
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: Strange RAID2 behavier...
2002-05-30 11:59 Strange RAID2 behavier me
@ 2002-05-30 12:37 ` me
2002-05-30 12:12 ` Lionel Bouton
0 siblings, 1 reply; 8+ messages in thread
From: me @ 2002-05-30 12:37 UTC (permalink / raw)
To: linux-kernel; +Cc: Andre Hedrick
I made the md2 a linear raid of one drive, now I can stop the md3.
This means that for example making md0 and md3 will make md3 unstoppable,
has this bug already been reported?
----- new raidtab -----
raiddev /dev/md0
raid-level 0
nr-raid-disks 2
persistent-superblock 1
chunk-size 4
device /dev/hde1
raid-disk 0
device /dev/hdi1
raid-disk 1
raiddev /dev/md1
raid-level 0
nr-raid-disks 2
persistent-superblock 1
chunk-size 4
device /dev/hdf1
raid-disk 0
device /dev/hdj1
raid-disk 1
raiddev /dev/md2
raid-level linear
nr-raid-disks 1
persistent-superblock 1
chunk-size 4
device /dev/hdk1
raid-disk 0
raiddev /dev/md3
raid-level 0
nr-raid-disks 2
persistent-superblock 1
chunk-size 4
device /dev/hdh1
raid-disk 0
device /dev/hdl1
raid-disk 1
-------------------------
On Thu, 30 May 2002 me@vger.org wrote:
> Ive got strange raid2 behaiver happening, ive got this raidtab:
>
> ----- raidtab -----
> raiddev /dev/md0
> raid-level 0
> nr-raid-disks 2
> persistent-superblock 1
> chunk-size 8
> device /dev/hde1
> raid-disk 0
> device /dev/hdi1
> raid-disk 1
>
> raiddev /dev/md1
> raid-level 0
> nr-raid-disks 2
> persistent-superblock 1
> chunk-size 8
> device /dev/hdf1
> raid-disk 0
> device /dev/hdj1
> raid-disk 1
>
> # not in use right now
> #raiddev /dev/md2
> # raid-level 0
> # nr-raid-disks 2
> # persistent-superblock 1
> # chunk-size 8
> # device /dev/hdg1
> # raid-disk 0
> # device /dev/hdk1
> # raid-disk 1
>
> raiddev /dev/md3
> raid-level 0
> nr-raid-disks 2
> persistent-superblock 1
> chunk-size 8
> device /dev/hdh1
> raid-disk 0
> device /dev/hdl1
> raid-disk 1
> ----- raidtab -----
>
> Im missing one disk on the md2 so im not using that right now but that
> should not affect the raid2 or make this happen:
>
> # raidstop --all /dev/md3
> /dev/md3: Device or resource busy
>
> # cat /proc/mdstat
> Personalities : [linear] [raid0] [raid1] [raid5] [multipath]
> read_ahead 1024 sectors
> md3 : active raid0 hdl1[1] hdh1[0]
> 240688000 blocks 8k chunks
>
> unused devices: <none>
>
> md0 and md1 gets turned off just fine, but not md3 ????
>
> I will try and rebuild the raid and mode md3 disks to md2 and see what
> happens.
>
> # uname -a
> Linux odd 2.4.19-pre9 #1 SMP Wed May 29 08:46:00 CEST 2002 i686 unknown
>
> There is no new 2.4.19-pre patch out from what i can find.
>
> Thanks in advance for the help.
>
>
>
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: Strange RAID2 behavier...
2002-05-30 12:12 ` Lionel Bouton
2002-05-30 12:21 ` Lionel Bouton
@ 2002-05-30 13:12 ` me
2002-05-30 15:01 ` me
1 sibling, 1 reply; 8+ messages in thread
From: me @ 2002-05-30 13:12 UTC (permalink / raw)
To: Lionel Bouton; +Cc: linux-kernel, Neil Brown
On Thu, 30 May 2002, Lionel Bouton wrote:
> On jeu, mai 30, 2002 at 02:37:53 +0200, me@vger.org wrote:
> > I made the md2 a linear raid of one drive, now I can stop the md3.
> > This means that for example making md0 and md3 will make md3 unstoppable,
> > has this bug already been reported?
>
> This could be a bug/constraint in raidtools as said in mad raidtab:
> "the parsing code isn't overly bright".
>
> I've had small glitches when using comments in raidtab. Try removing all
> your commented /dev/md2 related lines instead of building a dumb array.
>
Ok, i tought id fixed the problem (parsing problem) but it doesnt seam to
be that.
The md's ive got I later make a linear lvm partition from, may have
something to do with that.
I can start the raid, get lvm running, mount the lvm partition.
And then I umount the lvm partition and do vgchange -an to deactivate all
lvms. Now i can stop md0-1 but not md2-3.
--- strace raidstop ---
# strace raidstop /dev/md3
execve("/sbin/raidstop", ["raidstop", "/dev/md3"], [/* 14 vars */]) = 0
uname({sys="Linux", node="odd", ...}) = 0
brk(0) = 0x804fd50
open("/etc/ld.so.preload", O_RDONLY) = -1 ENOENT (No such file or
directory)
open("/etc/ld.so.cache", O_RDONLY) = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=9361, ...}) = 0
old_mmap(NULL, 9361, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40014000
close(3) = 0
open("/lib/libpopt.so.0", O_RDONLY) = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0`\21\0\000"...,
1024) = 1024
fstat64(3, {st_mode=S_IFREG|0644, st_size=20704, ...}) = 0
old_mmap(NULL, 23828, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40017000
mprotect(0x4001c000, 3348, PROT_NONE) = 0
old_mmap(0x4001c000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3,
0x4000) = 0x4001c000
close(3) = 0
open("/lib/libc.so.6", O_RDONLY) = 3
read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\30\222"...,
1024) = 1024
fstat64(3, {st_mode=S_IFREG|0755, st_size=1153784, ...}) = 0
old_mmap(NULL, 1166560, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) =
0x4001d000
mprotect(0x40130000, 40160, PROT_NONE) = 0
old_mmap(0x40130000, 24576, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED,
3, 0x113000) = 0x40130000
old_mmap(0x40136000, 15584, PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x40136000
close(3) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1,
0) = 0x4013a000
munmap(0x40014000, 9361) = 0
brk(0) = 0x804fd50
brk(0x804fef8) = 0x804fef8
brk(0x8050000) = 0x8050000
brk(0x8051000) = 0x8051000
open("/etc/raidtab", O_RDONLY) = 3
open("/dev/md0", O_RDONLY) = 4
ioctl(4, 0x800c0910, 0x804fd3c) = 0
close(4) = 0
open("/dev/md3", O_RDWR) = 4
fstat64(4, {st_mode=S_IFBLK|0660, st_rdev=makedev(9, 3), ...}) = 0
ioctl(4, 0x932, 0) = -1 EBUSY (Device or resource
busy)
dup(2) = 5
fcntl64(5, F_GETFL) = 0x8002 (flags
O_RDWR|O_LARGEFILE)
fstat64(5, {st_mode=S_IFCHR|0600, st_rdev=makedev(136, 0), ...}) = 0
old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1,
0) = 0x40014000
_llseek(5, 0, 0xbffff964, SEEK_CUR) = -1 ESPIPE (Illegal seek)
write(5, "/dev/md3: Device or resource bus"..., 34/dev/md3: Device or
resource busy
) = 34
close(5) = 0
munmap(0x40014000, 4096) = 0
_exit(1) = ?
------------------
Can this be a version problem with raidtool2 ? Im using debian testing
dist.
Any way i can check why/who has that device? lsof shows no md3.
Im clueless right now, sure the raid work even if md2 md3 are in use under
reboot but this not correct.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: Strange RAID2 behavier...
2002-05-30 13:12 ` me
@ 2002-05-30 15:01 ` me
0 siblings, 0 replies; 8+ messages in thread
From: me @ 2002-05-30 15:01 UTC (permalink / raw)
To: linux-lvm, lvm-devel; +Cc: Lionel Bouton, linux-kernel, Neil Brown
No worry linux-kernel, this seams to be a lvm problem.
Hi linux-lvm,
In short, I got 4 md devices (md0-4) that I make one device with using
lvm. When I'm going to shutdown this device it seams that lvm locks md2-3.
Booting linux without lvm there is no problem what so ever to start and
stop the raid mds.
Also as you read the futher down, the reason I could stop md3 was because
I hadnt used lvm yet that time and hadnt made any VGs or LVs or mounted
them. Also, the way i stop lvm is as described in the documents "umount,
vgchange -an".
Im using lvm-1.1rc-2, 2.4.19-pre9 and debian testing dist.
Im going to try running lvm as module and unloading the module after
stopping the lvm VGs and see if it still locks the mds.
If you need more info, please send a mail.
On Thu, 30 May 2002 me@vger.org wrote:
> On Thu, 30 May 2002, Lionel Bouton wrote:
>
> > On jeu, mai 30, 2002 at 02:37:53 +0200, me@vger.org wrote:
> > > I made the md2 a linear raid of one drive, now I can stop the md3.
> > > This means that for example making md0 and md3 will make md3 unstoppable,
> > > has this bug already been reported?
> >
> > This could be a bug/constraint in raidtools as said in mad raidtab:
> > "the parsing code isn't overly bright".
> >
> > I've had small glitches when using comments in raidtab. Try removing all
> > your commented /dev/md2 related lines instead of building a dumb array.
> >
>
> Ok, i tought id fixed the problem (parsing problem) but it doesnt seam to
> be that.
>
> The md's ive got I later make a linear lvm partition from, may have
> something to do with that.
>
> I can start the raid, get lvm running, mount the lvm partition.
> And then I umount the lvm partition and do vgchange -an to deactivate all
> lvms. Now i can stop md0-1 but not md2-3.
>
> --- strace raidstop ---
> # strace raidstop /dev/md3
> execve("/sbin/raidstop", ["raidstop", "/dev/md3"], [/* 14 vars */]) = 0
> uname({sys="Linux", node="odd", ...}) = 0
> brk(0) = 0x804fd50
> open("/etc/ld.so.preload", O_RDONLY) = -1 ENOENT (No such file or
> directory)
> open("/etc/ld.so.cache", O_RDONLY) = 3
> fstat64(3, {st_mode=S_IFREG|0644, st_size=9361, ...}) = 0
> old_mmap(NULL, 9361, PROT_READ, MAP_PRIVATE, 3, 0) = 0x40014000
> close(3) = 0
> open("/lib/libpopt.so.0", O_RDONLY) = 3
> read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0`\21\0\000"...,
> 1024) = 1024
> fstat64(3, {st_mode=S_IFREG|0644, st_size=20704, ...}) = 0
> old_mmap(NULL, 23828, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x40017000
> mprotect(0x4001c000, 3348, PROT_NONE) = 0
> old_mmap(0x4001c000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3,
> 0x4000) = 0x4001c000
> close(3) = 0
> open("/lib/libc.so.6", O_RDONLY) = 3
> read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\30\222"...,
> 1024) = 1024
> fstat64(3, {st_mode=S_IFREG|0755, st_size=1153784, ...}) = 0
> old_mmap(NULL, 1166560, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) =
> 0x4001d000
> mprotect(0x40130000, 40160, PROT_NONE) = 0
> old_mmap(0x40130000, 24576, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED,
> 3, 0x113000) = 0x40130000
> old_mmap(0x40136000, 15584, PROT_READ|PROT_WRITE,
> MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x40136000
> close(3) = 0
> old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1,
> 0) = 0x4013a000
> munmap(0x40014000, 9361) = 0
> brk(0) = 0x804fd50
> brk(0x804fef8) = 0x804fef8
> brk(0x8050000) = 0x8050000
> brk(0x8051000) = 0x8051000
> open("/etc/raidtab", O_RDONLY) = 3
> open("/dev/md0", O_RDONLY) = 4
> ioctl(4, 0x800c0910, 0x804fd3c) = 0
> close(4) = 0
> open("/dev/md3", O_RDWR) = 4
> fstat64(4, {st_mode=S_IFBLK|0660, st_rdev=makedev(9, 3), ...}) = 0
> ioctl(4, 0x932, 0) = -1 EBUSY (Device or resource
> busy)
> dup(2) = 5
> fcntl64(5, F_GETFL) = 0x8002 (flags
> O_RDWR|O_LARGEFILE)
> fstat64(5, {st_mode=S_IFCHR|0600, st_rdev=makedev(136, 0), ...}) = 0
> old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1,
> 0) = 0x40014000
> _llseek(5, 0, 0xbffff964, SEEK_CUR) = -1 ESPIPE (Illegal seek)
> write(5, "/dev/md3: Device or resource bus"..., 34/dev/md3: Device or
> resource busy
> ) = 34
> close(5) = 0
> munmap(0x40014000, 4096) = 0
> _exit(1) = ?
> ------------------
>
> Can this be a version problem with raidtool2 ? Im using debian testing
> dist.
>
> Any way i can check why/who has that device? lsof shows no md3.
>
> Im clueless right now, sure the raid work even if md2 md3 are in use under
> reboot but this not correct.
>
>
>
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: Strange RAID2 behavier...
2002-05-30 12:01 Neil Brown
@ 2002-05-30 12:42 ` me
0 siblings, 0 replies; 8+ messages in thread
From: me @ 2002-05-30 12:42 UTC (permalink / raw)
To: Neil Brown; +Cc: linux-kernel, Andre Hedrick
On Thu, 30 May 2002, Neil Brown wrote:
> On Thursday May 30, me@vger.org wrote:
> >
> > # raidstop --all /dev/md3
> > /dev/md3: Device or resource busy
>
> This just means that /dev/md3 is busy.
> Is it mounted? Does any process have it open?
>
md3 is unmounted and not in use by anything.
> There is a bug in one version of raidtools that causes raidstop to
> incorrectly report this error, but I think that bug only affects
> /dev/md0..
> What does
> strace raidstop /dev/md3
> show?
>
cant do that now =) because ive already "fixed" the problem (read my other
mails).
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: Strange RAID2 behavier...
@ 2002-05-30 12:01 Neil Brown
2002-05-30 12:42 ` me
0 siblings, 1 reply; 8+ messages in thread
From: Neil Brown @ 2002-05-30 12:01 UTC (permalink / raw)
To: me; +Cc: linux-kernel, Andre Hedrick
On Thursday May 30, me@vger.org wrote:
>
> # raidstop --all /dev/md3
> /dev/md3: Device or resource busy
This just means that /dev/md3 is busy.
Is it mounted? Does any process have it open?
There is a bug in one version of raidtools that causes raidstop to
incorrectly report this error, but I think that bug only affects
/dev/md0..
What does
strace raidstop /dev/md3
show?
NeilBrown
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2002-05-30 14:24 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2002-05-30 11:59 Strange RAID2 behavier me
2002-05-30 12:37 ` me
2002-05-30 12:12 ` Lionel Bouton
2002-05-30 12:21 ` Lionel Bouton
2002-05-30 13:12 ` me
2002-05-30 15:01 ` me
2002-05-30 12:01 Neil Brown
2002-05-30 12:42 ` me
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®