Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
Hi Lars,
I changed the drbd.conf to:
on dlssi3 {
device=/dev/nbd/0
disk=/dev/hde2
address=15.70.191.200
port=7788
meta-disk=/dev/hde5
meta-index=0
}
where /dev/hde2 is 10G and /dev/hde5 is 1G. I'm attaching a file with
the output of starting the drbd device and the fsck failure. Also
attached is the drbd.conf. Note that the second node has not been
started at this point.
What am I doing wrong?
En Chiang
On Thu, 2004-03-25 at 22:02, Lars Ellenberg wrote:
> / 2004-03-25 20:09:41 +0530
> \ En Chiang Lee:
> > Hi,
> >
> > I was trying to mirror an existing ext3 filesystem using drbd to another
> > node in an OpenSSI cluster, so I associated /dev/nbd/0 with /dev/hde2.
> >
> > on dlssi3 {
> > device=/dev/nbd/0
> > disk=/dev/hde2
> > address=15.70.191.200
> > port=7788
> > meta-disk=internal
> ^^^^^^^^
>
> > meta-index=-1
> > }
> >
> > With this, I am able to mount the filesystem. However, when I try fsck
> > on /dev/nbd/0, I get the following error:
> >
> > [root at dlssi3 root]# fsck /dev/nbd/0
> > fsck 1.32 (09-Nov-2002)
> > e2fsck 1.32 (09-Nov-2002)
> > /: recovering journal
> > The filesystem size (according to the superblock) is 2560351 blocks
> > The physical size of the device is 2527583 blocks
> > Either the superblock or the partition table is likely to be corrupt!
> > Abort<y>? yes
>
> > (2560351-2527583) * (4K blocks) == 128 M
>
> you pretty likely already corrupted the underlying filesystem
> (at least the last 128M of it).
>
> If you want to have "meta-disk=internal",
> the right way to go is to create the filesystem on the /dev/nbd/0,
> then it will get the right size.
>
>
> > In the archives, there is mention of using "size=<somesize>", but that
> > is deprecated according the the sample drbd.conf. (If I was to use size=
> > how would the value be specified? fs blocks? GB?)
>
> the size parameter is in 1K, but also takes M and G as unit modifiers.
>
> > I'm trying this with drbd-0.7_pre5.
> >
> > How would I go about using drbd on an existing file system?
>
> You can have the meta data in some other device (not internal), and
> adjust the meta-disk and meta-index accordingly.
>
> OR you can resize (shrink) the file system first.
> *BEFORE* the last 128M of the device get overwritten with bitmap data.
>
> Lars Ellenberg
> _______________________________________________
> drbd-user mailing list
> drbd-user at lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user
>
-------------- next part --------------
[root at dlssi3 /]# cat /proc/drbd
version: 0.7-pre5 (api:72/proto:72)
0: cs:Unconfigured
1: cs:Unconfigured
[root at dlssi3 /]# mkfs -t ext3 /dev/hde2
mke2fs 1.32 (09-Nov-2002)
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
1281696 inodes, 2560359 blocks
128017 blocks (5.00%) reserved for the super user
First data block=0
79 block groups
32768 blocks per group, 32768 fragments per group
16224 inodes per group
Superblock backups stored on blocks:
32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632
Writing inode tables: done
Creating journal (8192 blocks): done
Writing superblocks and filesystem accounting information: done
This filesystem will be automatically checked every 21 mounts or
180 days, whichever comes first. Use tune2fs -c or -i to override.
[root at dlssi3 /]# drbdadm up r0
[root at dlssi3 /]# drbdadm primary r0
[root at dlssi3 /]# cat /proc/drbd
version: 0.7-pre5 (api:72/proto:72)
0: cs:WFConnection st:Primary/Unknown ld:Consistent
ns:0 nr:0 dw:0 dr:0 al:0 bm:2571 lo:0 pe:0 ua:0
1: cs:Unconfigured
[root at dlssi3 /]# fsck /dev/nbd/0
fsck 1.32 (09-Nov-2002)
e2fsck 1.32 (09-Nov-2002)
The filesystem size (according to the superblock) is 2560359 blocks
The physical size of the device is 320044 blocks
Either the superblock or the partition table is likely to be corrupt!
Abort<y>? yes
[root at dlssi3 /]#
[root at dlssi3 /]# blockdev --getsize /dev/nbd/0
2560358
[root at dlssi3 /]# blockdev --getsize /dev/hde2
20482875
[root at dlssi3 /]#
-------------- next part --------------
resource "r0" {
protocol = B
incon-degr-cmd="halt -f"
startup {
wfc-timeout=10 # wait for 10 seconds for connection otherwise continue
degr-wfc-timeout=10 # 2 minutes.
}
disk {
on-io-error = panic
}
syncer {
rate=10M
group=1
al-extents=257
}
on dlssi3 {
device=/dev/nbd/0
disk=/dev/hde2
address=15.70.191.200
port=7788
meta-disk=/dev/hde5
meta-index=0
}
on dlssi1 {
device=/dev/nbd/0
disk=/dev/ida/c0d0p3
address=15.70.191.198
port=7788
meta-disk=internal
meta-index=-1
}
}