Mailing List Archive

All I can say is, yes, you're asking for trouble with this setup, if my
latest experience is worth anything.

XFS doesn't work on drbd yet, either. Maybe ext2 will?

If you can bear it, go back to the 2.2.x kernel where drbd worked like a
champ.

See the archives for what happened to me last week.

Cheers,
Dan


--
Dan Yocum
Sloan Digital Sky Survey, Fermilab 630.840.6509
yocum@example.com, http://www.sdss.org
SDSS. Mapping the Universe.

* Rob Ostrander <ostrander@example.com> [011219 15:48]:
> Hello
> I'm considering putting together a ha-nfs system. Based on skimming the
> posts to the lists over the last month or so I'm a little concerned with
> the system I had in mind. I was planning on using Red Hat 7.2 on whatever
> 2.4.x recommended kernel, with a ext3 file system, and possibly md/raid 5
> for each machine(with hd not nbd). Please let me know if I'm asking for
> trouble, being excessive etc...
> Rob Ostrander
> Integrated Sensors Inc.
> http://www.sensors.com
> phone: 315-798-1377
> fax: 315-798-8950
>

I would say a good choice. I would suggest 2.4.latest and drbd-0.6.1-latest.
While I have good experience with ext2 I am goint to test ext3 on DRBD.
As long as you do not need reiserf's advanced features would suggest.

About the list...

I guess people with working systems do not need to write about that fact
on the list, while people with problems...


-Philipp
--
: Dipl-Ing Philipp Reisner Tel +43-1-8974897-750 :
: LINBIT Information Technologies GmbH http://www.linbit.com :
: Sechshauserstr 48, 1150 Wien :

Indeed, people with working systems don't need to post to the list, but I
suppose we should every now and then. We're using an older kernel, and an
older drbd, but that's only because we've been working on our uptime now
for a while. When we set things up, the list traffic looked just the same
as it does today... There will always be people having troubles getting
things to work right for various reasons. We had troubles too on our
initial setup. The stability problems turned out to be a funky motherboard
in the end, but I ended up ripping quite a bit of hair out before we
figured that one out ;).

Our uptime isn't so impressive right now, but that's because of our 5
hours of downtime that we incurred when we moved our offices 20 miles.
Before our move, we had an additional 200+ days of uptime. So far, things
have worked absolutely perfectly.

I guess now would be a good time to say thanks for creating such a
wonderful piece of software. My boss sleeps at night because of drbd.

-Jeff Buck (jeffb@example.com)

----------------------------------------

[root@jupiter /root]# uptime
9:38am up 105 days, 8:00, 2 users, load average: 0.00, 0.00, 0.00

----------------------------------------

[root@jupiter /root]# cat /proc/drbd
version : 58

0: cs:Connected st:Primary/Secondary ns:309813136 nr:0 dw:77453284
dr:144238985 gc:6,3346,5

-----------------------------------------

[root@jupiter /root]# df
Filesystem Size Used Avail Use% Mounted on
/dev/rd/c0d0p2 5.8G 4.7G 743M 87% /
/dev/rd/c0d0p1 45M 19M 24M 45% /boot
/dev/nb0 178G 62G 106G 37% /data
/dev/hda1 14G 20k 13G 0% /mnt/hda1

---------------------------------------

On Thu, 20 Dec 2001, Philipp Reisner wrote:

> * Rob Ostrander <ostrander@example.com> [011219 15:48]:
> > Hello
> > I'm considering putting together a ha-nfs system. Based on skimming the
> > posts to the lists over the last month or so I'm a little concerned with
> > the system I had in mind. I was planning on using Red Hat 7.2 on whatever
>
> About the list...
>
> I guess people with working systems do not need to write about that fact
> on the list, while people with problems...

Generally, I'd agree. drbd .58 on 2.2.18 with ext2 and no software raid
worked great. I couldn't break that thing if I tried, but now, something
has gone very wrong, and I don't know where. drbd 61pre6 on 2.4.9, linear
software raid and ext3 was a very, very bad thing and I almost lost my shirt
- thank god for backups.

I sent my error messages, but got no response. Others have the same
problem, so I'm not the only one. We're starting to think that it might
have something to do with running drbd on software RAID - I wouldn't be
surprised.

Does anyone have drbd working reliably on a 2.4.x kernel using a software
raid array and ext3?

Cheers,
Dan

Jeff Buck wrote:
>
> Indeed, people with working systems don't need to post to the list, but I
> suppose we should every now and then. We're using an older kernel, and an
> older drbd, but that's only because we've been working on our uptime now
> for a while. When we set things up, the list traffic looked just the same
> as it does today... There will always be people having troubles getting
> things to work right for various reasons. We had troubles too on our
> initial setup. The stability problems turned out to be a funky motherboard
> in the end, but I ended up ripping quite a bit of hair out before we
> figured that one out ;).
>
> Our uptime isn't so impressive right now, but that's because of our 5
> hours of downtime that we incurred when we moved our offices 20 miles.
> Before our move, we had an additional 200+ days of uptime. So far, things
> have worked absolutely perfectly.
>
> I guess now would be a good time to say thanks for creating such a
> wonderful piece of software. My boss sleeps at night because of drbd.
>
> -Jeff Buck (jeffb@example.com)
>
> ----------------------------------------
>
> [root@jupiter /root]# uptime
> 9:38am up 105 days, 8:00, 2 users, load average: 0.00, 0.00, 0.00
>
> ----------------------------------------
>
> [root@jupiter /root]# cat /proc/drbd
> version : 58
>
> 0: cs:Connected st:Primary/Secondary ns:309813136 nr:0 dw:77453284
> dr:144238985 gc:6,3346,5
>
> -----------------------------------------
>
> [root@jupiter /root]# df
> Filesystem Size Used Avail Use% Mounted on
> /dev/rd/c0d0p2 5.8G 4.7G 743M 87% /
> /dev/rd/c0d0p1 45M 19M 24M 45% /boot
> /dev/nb0 178G 62G 106G 37% /data
> /dev/hda1 14G 20k 13G 0% /mnt/hda1
>
> ---------------------------------------
>
> On Thu, 20 Dec 2001, Philipp Reisner wrote:
>
> > * Rob Ostrander <ostrander@example.com> [011219 15:48]:
> > > Hello
> > > I'm considering putting together a ha-nfs system. Based on skimming the
> > > posts to the lists over the last month or so I'm a little concerned with
> > > the system I had in mind. I was planning on using Red Hat 7.2 on whatever
> >
> > About the list...
> >
> > I guess people with working systems do not need to write about that fact
> > on the list, while people with problems...
>
> _______________________________________________
> DRBD-devel mailing list
> DRBD-devel@example.com
> https://lists.sourceforge.net/lists/listinfo/drbd-devel



--
Dan Yocum
Sloan Digital Sky Survey, Fermilab 630.840.6509
yocum@example.com, http://www.sdss.org
SDSS. Mapping the Universe.

Hi

I have a cluster working with two SMP machines with kernel 2.4.9. and
drdb-0.6.1-pre5. and I had several problems, and especially one in which I almost
lost all my data (the tape was also broken). But McGuyver recovered everything!
Now I'm using in another test platform (also with one SMP machine) kernel 2.4.14 and
drbd-0.6.1.pre7 and it seems much more reliable. I meam much, much, more realiable.

Best regards
Raul

Dan Yocum wrote:

> Generally, I'd agree. drbd .58 on 2.2.18 with ext2 and no software raid
> worked great. I couldn't break that thing if I tried, but now, something
> has gone very wrong, and I don't know where. drbd 61pre6 on 2.4.9, linear
> software raid and ext3 was a very, very bad thing and I almost lost my shirt
> - thank god for backups.
>
> I sent my error messages, but got no response. Others have the same
> problem, so I'm not the only one. We're starting to think that it might
> have something to do with running drbd on software RAID - I wouldn't be
> surprised.
>
> Does anyone have drbd working reliably on a 2.4.x kernel using a software
> raid array and ext3?
>
> Cheers,
> Dan
>
> Jeff Buck wrote:
> >
> > Indeed, people with working systems don't need to post to the list, but I
> > suppose we should every now and then. We're using an older kernel, and an
> > older drbd, but that's only because we've been working on our uptime now
> > for a while. When we set things up, the list traffic looked just the same
> > as it does today... There will always be people having troubles getting
> > things to work right for various reasons. We had troubles too on our
> > initial setup. The stability problems turned out to be a funky motherboard
> > in the end, but I ended up ripping quite a bit of hair out before we
> > figured that one out ;).
> >
> > Our uptime isn't so impressive right now, but that's because of our 5
> > hours of downtime that we incurred when we moved our offices 20 miles.
> > Before our move, we had an additional 200+ days of uptime. So far, things
> > have worked absolutely perfectly.
> >
> > I guess now would be a good time to say thanks for creating such a
> > wonderful piece of software. My boss sleeps at night because of drbd.
> >
> > -Jeff Buck (jeffb@example.com)
> >
> > ----------------------------------------
> >
> > [root@jupiter /root]# uptime
> > 9:38am up 105 days, 8:00, 2 users, load average: 0.00, 0.00, 0.00
> >
> > ----------------------------------------
> >
> > [root@jupiter /root]# cat /proc/drbd
> > version : 58
> >
> > 0: cs:Connected st:Primary/Secondary ns:309813136 nr:0 dw:77453284
> > dr:144238985 gc:6,3346,5
> >
> > -----------------------------------------
> >
> > [root@jupiter /root]# df
> > Filesystem Size Used Avail Use% Mounted on
> > /dev/rd/c0d0p2 5.8G 4.7G 743M 87% /
> > /dev/rd/c0d0p1 45M 19M 24M 45% /boot
> > /dev/nb0 178G 62G 106G 37% /data
> > /dev/hda1 14G 20k 13G 0% /mnt/hda1
> >
> > ---------------------------------------
> >
> > On Thu, 20 Dec 2001, Philipp Reisner wrote:
> >
> > > * Rob Ostrander <ostrander@example.com> [011219 15:48]:
> > > > Hello
> > > > I'm considering putting together a ha-nfs system. Based on skimming the
> > > > posts to the lists over the last month or so I'm a little concerned with
> > > > the system I had in mind. I was planning on using Red Hat 7.2 on whatever
> > >
> > > About the list...
> > >
> > > I guess people with working systems do not need to write about that fact
> > > on the list, while people with problems...
> >
> > _______________________________________________
> > DRBD-devel mailing list
> > DRBD-devel@example.com
> > https://lists.sourceforge.net/lists/listinfo/drbd-devel
>
> --
> Dan Yocum
> Sloan Digital Sky Survey, Fermilab 630.840.6509
> yocum@example.com, http://www.sdss.org
> SDSS. Mapping the Universe.
>
> _______________________________________________
> DRBD-devel mailing list
> DRBD-devel@example.com
> https://lists.sourceforge.net/lists/listinfo/drbd-devel

Raul,

What filesystem are you using and how big is the drbd volume?

Are you using any software RAID?

Thanks,
Dan


Raul Oliveira wrote:
>
> Hi
>
> I have a cluster working with two SMP machines with kernel 2.4.9. and
> drdb-0.6.1-pre5. and I had several problems, and especially one in which I almost
> lost all my data (the tape was also broken). But McGuyver recovered everything!
> Now I'm using in another test platform (also with one SMP machine) kernel 2.4.14 and
> drbd-0.6.1.pre7 and it seems much more reliable. I meam much, much, more realiable.
>
> Best regards
> Raul
>
> Dan Yocum wrote:
>
> > Generally, I'd agree. drbd .58 on 2.2.18 with ext2 and no software raid
> > worked great. I couldn't break that thing if I tried, but now, something
> > has gone very wrong, and I don't know where. drbd 61pre6 on 2.4.9, linear
> > software raid and ext3 was a very, very bad thing and I almost lost my shirt
> > - thank god for backups.
> >
> > I sent my error messages, but got no response. Others have the same
> > problem, so I'm not the only one. We're starting to think that it might
> > have something to do with running drbd on software RAID - I wouldn't be
> > surprised.
> >
> > Does anyone have drbd working reliably on a 2.4.x kernel using a software
> > raid array and ext3?
> >
> > Cheers,
> > Dan
> >
> > Jeff Buck wrote:
> > >
> > > Indeed, people with working systems don't need to post to the list, but I
> > > suppose we should every now and then. We're using an older kernel, and an
> > > older drbd, but that's only because we've been working on our uptime now
> > > for a while. When we set things up, the list traffic looked just the same
> > > as it does today... There will always be people having troubles getting
> > > things to work right for various reasons. We had troubles too on our
> > > initial setup. The stability problems turned out to be a funky motherboard
> > > in the end, but I ended up ripping quite a bit of hair out before we
> > > figured that one out ;).
> > >
> > > Our uptime isn't so impressive right now, but that's because of our 5
> > > hours of downtime that we incurred when we moved our offices 20 miles.
> > > Before our move, we had an additional 200+ days of uptime. So far, things
> > > have worked absolutely perfectly.
> > >
> > > I guess now would be a good time to say thanks for creating such a
> > > wonderful piece of software. My boss sleeps at night because of drbd.
> > >
> > > -Jeff Buck (jeffb@example.com)
> > >
> > > ----------------------------------------
> > >
> > > [root@jupiter /root]# uptime
> > > 9:38am up 105 days, 8:00, 2 users, load average: 0.00, 0.00, 0.00
> > >
> > > ----------------------------------------
> > >
> > > [root@jupiter /root]# cat /proc/drbd
> > > version : 58
> > >
> > > 0: cs:Connected st:Primary/Secondary ns:309813136 nr:0 dw:77453284
> > > dr:144238985 gc:6,3346,5
> > >
> > > -----------------------------------------
> > >
> > > [root@jupiter /root]# df
> > > Filesystem Size Used Avail Use% Mounted on
> > > /dev/rd/c0d0p2 5.8G 4.7G 743M 87% /
> > > /dev/rd/c0d0p1 45M 19M 24M 45% /boot
> > > /dev/nb0 178G 62G 106G 37% /data
> > > /dev/hda1 14G 20k 13G 0% /mnt/hda1
> > >
> > > ---------------------------------------
> > >
> > > On Thu, 20 Dec 2001, Philipp Reisner wrote:
> > >
> > > > * Rob Ostrander <ostrander@example.com> [011219 15:48]:
> > > > > Hello
> > > > > I'm considering putting together a ha-nfs system. Based on skimming the
> > > > > posts to the lists over the last month or so I'm a little concerned with
> > > > > the system I had in mind. I was planning on using Red Hat 7.2 on whatever
> > > >
> > > > About the list...
> > > >
> > > > I guess people with working systems do not need to write about that fact
> > > > on the list, while people with problems...
> > >
> > > _______________________________________________
> > > DRBD-devel mailing list
> > > DRBD-devel@example.com
> > > https://lists.sourceforge.net/lists/listinfo/drbd-devel
> >
> > --
> > Dan Yocum
> > Sloan Digital Sky Survey, Fermilab 630.840.6509
> > yocum@example.com, http://www.sdss.org
> > SDSS. Mapping the Universe.
> >
> > _______________________________________________
> > DRBD-devel mailing list
> > DRBD-devel@example.com
> > https://lists.sourceforge.net/lists/listinfo/drbd-devel



--
Dan Yocum
Sloan Digital Sky Survey, Fermilab 630.840.6509
yocum@example.com, http://www.sdss.org
SDSS. Mapping the Universe.

Hi Dan

> What filesystem are you using and how big is the drbd volume?

I'm using ext2 with LVM. I tried reiserfs but I got so many errors with 2.4.9,
LVM and drbd that I give up.
I have two drbd volumes, one with 10GB and the other with 25GB.

> Are you using any software RAID?
No.

Right now in my test platform I'using everything similar to the first cluster,
and the only difference is that the drbd volumes are quite small 1GB.
I'm planning to move the situation in the test platform to real cluster this
Friday.

Best regards
Raul