Lustre discuss - Feb 2008 - obdfilter/datafs-OST0000/recovery_status

I''m evaluating lustre. I''m trying what I think is a
basic/simple
ethernet config. with MDT and OST on the same node. Can someone tell
me if the following (~150 second recovery occurring when small 190 GB
OST is re-mounted) is expected behavior or if I''m missing something?
I thought I would send this and continue with the eval while awaiting
a
response.
I''m using lustre release 1.6.4.2 with the vanilla 2.6.18.8 kernel with
a Scientific Linux 5 (derived from RHEL5) distro with e2fsprogs
1.40.4.cfs1. I''m doing the following:
aaa()
{
  set -x
  dmesg -c >/dev/null
  mkfs.lustre --fsname datafs --mdt --mgs --reformat /dev/sda1
  mkfs.lustre --fsname datafs --ost --mgsnode=pool4 at tcp --reformat /
dev/sda2
  e2label /dev/sda1
  e2label /dev/sda2
  mount.lustre /dev/sda1 /mnt/data/mdt
  mount.lustre /dev/sda2 /mnt/data/ost0
  dmesg -c >dmesg.0
  mount.lustre pool4 at tcp:/datafs /mnt/datafs
  dmesg -c >dmesg.1
  umount /mnt/datafs
  umount /mnt/data/ost0
  umount /mnt/data/mdt
  e2label /dev/sda1
  e2label /dev/sda2
  dmesg -c >dmesg.2
  mount.lustre /dev/sda1 /mnt/data/mdt
  mount.lustre /dev/sda2 /mnt/data/ost0
  dmesg -c >dmesg.3
  while cat /proc/fs/lustre/obdfilter/datafs-OST0000/recovery_status \
        | egrep ''RECOVERING|time remaining'';do sleep 30;done
  mount.lustre pool4 at tcp:/datafs /mnt/datafs
}
aaa 2>&1 | tee aaa.0; dmesg -c >dmesg.4

The files dmesg.{0,1,2,3,4} and aaa.0 are available at:
http://fnapcf.fnal.gov/~ron/lustre-1.6.4.2-dmesg3-e2fsprog/

Here is aaa.0 which shows the recovery:
+ dmesg -c
+ mkfs.lustre --fsname datafs --mdt --mgs --reformat /dev/sda1
WARNING: MDS group upcall is not set, use ''NONE''

   Permanent disk data:
Target:     datafs-MDTffff
Index:      unassigned
Lustre FS:  datafs
Mount type: ldiskfs
Flags:      0x75
              (MDT MGS needs_index first_time update )
Persistent mount opts: errors=remount-ro,iopen_nopriv,user_xattr
Parameters:

device size = 95367MB
formatting backing filesystem ldiskfs on /dev/sda1
        target name  datafs-MDTffff
        4k blocks     0
        options        -J size=400 -i 4096 -I 512 -q -O dir_index -F
mkfs_cmd = mkfs.ext2 -j -b 4096 -L datafs-MDTffff  -J size=400 -i 4096
-I 512 -q -O dir_index -F /dev/sda1
Writing CONFIGS/mountdata
+ mkfs.lustre --fsname datafs --ost --mgsnode=pool4 at tcp --reformat /
dev/sda2

   Permanent disk data:
Target:     datafs-OSTffff
Index:      unassigned
Lustre FS:  datafs
Mount type: ldiskfs
Flags:      0x72
              (OST needs_index first_time update )
Persistent mount opts: errors=remount-ro,extents,mballoc
Parameters: mgsnode=192.168.241.247 at tcp

device size = 190734MB
formatting backing filesystem ldiskfs on /dev/sda2
        target name  datafs-OSTffff
        4k blocks     0
        options        -J size=400 -i 16384 -I 256 -q -O dir_index -F
mkfs_cmd = mkfs.ext2 -j -b 4096 -L datafs-OSTffff  -J size=400 -i
16384 -I 256 -q -O dir_index -F /dev/sda2
Writing CONFIGS/mountdata
+ e2label /dev/sda1
datafs-MDTffff
+ e2label /dev/sda2
datafs-OSTffff
+ mount.lustre /dev/sda1 /mnt/data/mdt
+ mount.lustre /dev/sda2 /mnt/data/ost0
+ dmesg -c
+ mount.lustre pool4 at tcp:/datafs /mnt/datafs
+ dmesg -c
+ umount /mnt/datafs
+ umount /mnt/data/ost0
+ umount /mnt/data/mdt
+ e2label /dev/sda1
datafs-MDT0000
+ e2label /dev/sda2
datafs-OST0000
+ dmesg -c
+ mount.lustre /dev/sda1 /mnt/data/mdt
+ mount.lustre /dev/sda2 /mnt/data/ost0
+ dmesg -c
+ cat /proc/fs/lustre/obdfilter/datafs-OST0000/recovery_status
+ egrep ''RECOVERING|time remaining''
status: RECOVERING
time remaining: 250
+ sleep 30
+ cat /proc/fs/lustre/obdfilter/datafs-OST0000/recovery_status
+ egrep ''RECOVERING|time remaining''
status: RECOVERING
time remaining: 245
+ sleep 30
+ cat /proc/fs/lustre/obdfilter/datafs-OST0000/recovery_status
+ egrep ''RECOVERING|time remaining''
status: RECOVERING
time remaining: 215
+ sleep 30
+ cat /proc/fs/lustre/obdfilter/datafs-OST0000/recovery_status
+ egrep ''RECOVERING|time remaining''
status: RECOVERING
time remaining: 185
+ sleep 30
+ cat /proc/fs/lustre/obdfilter/datafs-OST0000/recovery_status
+ egrep ''RECOVERING|time rg''
status: RECOVERING
time remaining: 5
+ sleep 30
+ cat /proc/fs/lustre/obdfilter/datafs-OST0000/recovery_status
+ egrep ''RECOVERING|time remaining''
+ mount.lustre pool4 at tcp:/datafs /mnt/datafs


Thanks,
Ron

On Tue, 2008-02-05 at 11:41 -0800, Ron wrote:
> I''m evaluating lustre. I''m trying what I think is a
basic/simple
> ethernet config. with MDT and OST on the same node. Can someone tell
> me if the following (~150 second recovery occurring when small 190 GB
> OST is re-mounted) is expected behavior or if I''m missing
something?
it''s expected given that you mounted the MDT first.  There is nothing
technically wrong with this (mountconf allows you to mount in any order)
except that when the OSTs mount, recovery will be needed.  To avoid
this, mount the OSTs first.

b.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url :
http://lists.lustre.org/pipermail/lustre-discuss/attachments/20080205/036d8fc8/attachment-0002.bin

On Feb 05, 2008  15:19 -0500, Brian J. Murrell wrote:
> On Tue, 2008-02-05 at 11:41 -0800, Ron wrote:
> > I''m evaluating lustre. I''m trying what I think is a
basic/simple
> > ethernet config. with MDT and OST on the same node. Can someone tell
> > me if the following (~150 second recovery occurring when small 190 GB
> > OST is re-mounted) is expected behavior or if I''m missing
something?
> 
> it''s expected given that you mounted the MDT first.  There is
nothing
> technically wrong with this (mountconf allows you to mount in any order)
> except that when the OSTs mount, recovery will be needed.  To avoid
> this, mount the OSTs first.
More importantly, _unmount_ the MDT first, since it is a "client" of
the
OSTs, and the OSTs are waiting for this client to reconnect when they
restart.

Cheers, Andreas
--
Andreas Dilger
Sr. Staff Engineer, Lustre Group
Sun Microsystems of Canada, Inc.