Mailing List Archive

BarrierAck errors on my DRBD
Hi,

After fixing an other error I started to see the following error on my DRBD:

kernel: block drbd0: BAD! BarrierAck #2778697761 received, expected
#2778697760!

Can anybody explain what the BarrierAck error means?
What is the cause of the problem?
How to solve the problem?

Thanks for any hints.

--
Dr. Michael Schwartzkopff
Guardinistr. 63
81375 München

Tel: (0163) 172 50 98
Fax: (089) 620 304 13
Re: BarrierAck errors on my DRBD [ In reply to ]
On Thu, Sep 15, 2011 at 10:29:47AM +0200, Michael Schwartzkopff wrote:
> Hi,
>
> After fixing an other error I started to see the following error on my DRBD:
>
> kernel: block drbd0: BAD! BarrierAck #2778697761 received, expected
> #2778697760!
>
> Can anybody explain what the BarrierAck error means?
> What is the cause of the problem?
> How to solve the problem?

DRBD version, kernel version, what "other error",
what did you do to trigger this,
can you easily reproduce, etc.

Most likely: Upgrade.



--
: Lars Ellenberg
: LINBIT | Your Way to High Availability
: DRBD/HA support and consulting http://www.linbit.com
_______________________________________________
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user
Re: BarrierAck errors on my DRBD [ In reply to ]
> On Thu, Sep 15, 2011 at 10:29:47AM +0200, Michael Schwartzkopff wrote:
> > Hi,
> >
> > After fixing an other error I started to see the following error on my
> > DRBD:
> >
> > kernel: block drbd0: BAD! BarrierAck #2778697761 received, expected
> > #2778697760!
> >
> > Can anybody explain what the BarrierAck error means?
> > What is the cause of the problem?
> > How to solve the problem?
>
> DRBD version,

8.3.10

> kernel version,

2.6.39.3

> what "other error",

Since we have a quite limited test setup with virtual machines I saw lots of

kernel: block drbd0: PingAck did not arrive in time.

I solved that problem
- limiting sync to 1M
- setting ping-int to 30 (seconds)
- setting ping-timeout to 20 (1/10th of seconds, i.e. 2 sec)

> what did you do to trigger this,

Nothing. It happens about twice per hour.

> can you easily reproduce, etc.

Yes. That is the annoying problem.

> Most likely: Upgrade.

From kernel 2.6.39 to?

Greetings,

--
Dr. Michael Schwartzkopff
Guardinistr. 63
81375 München

Tel: (0163) 172 50 98
Fax: (089) 620 304 13
Re: BarrierAck errors on my DRBD [ In reply to ]
On Thu, Sep 15, 2011 at 11:14:53AM +0200, Michael Schwartzkopff wrote:
> > On Thu, Sep 15, 2011 at 10:29:47AM +0200, Michael Schwartzkopff wrote:
> > > Hi,
> > >
> > > After fixing an other error I started to see the following error on my
> > > DRBD:
> > >
> > > kernel: block drbd0: BAD! BarrierAck #2778697761 received, expected
> > > #2778697760!
> > >
> > > Can anybody explain what the BarrierAck error means?
> > > What is the cause of the problem?
> > > How to solve the problem?
> >
> > DRBD version,
>
> 8.3.10

The in-tree or the out-of-tree thing?

> > kernel version,
>
> 2.6.39.3
>
> > what "other error",
>
> Since we have a quite limited test setup with virtual machines I saw lots of
>
> kernel: block drbd0: PingAck did not arrive in time.
>
> I solved that problem
> - limiting sync to 1M
> - setting ping-int to 30 (seconds)
> - setting ping-timeout to 20 (1/10th of seconds, i.e. 2 sec)
>
> > what did you do to trigger this,
>
> Nothing. It happens about twice per hour.
>
> > can you easily reproduce, etc.
>
> Yes. That is the annoying problem.
>
> > Most likely: Upgrade.
>
> From kernel 2.6.39 to?

From 8.3.10 in-kernel (I assume)
to 8.3.11-3 out-of-tree?


--
: Lars Ellenberg
: LINBIT | Your Way to High Availability
: DRBD/HA support and consulting http://www.linbit.com
_______________________________________________
drbd-user mailing list
drbd-user@lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user
Re: BarrierAck errors on my DRBD [ In reply to ]
> On Thu, Sep 15, 2011 at 11:14:53AM +0200, Michael Schwartzkopff wrote:
> > > On Thu, Sep 15, 2011 at 10:29:47AM +0200, Michael Schwartzkopff wrote:
> > > > Hi,
> > > >
> > > > After fixing an other error I started to see the following error on
> > > > my DRBD:
> > > >
> > > > kernel: block drbd0: BAD! BarrierAck #2778697761 received, expected
> > > > #2778697760!
> > > >
> > > > Can anybody explain what the BarrierAck error means?
> > > > What is the cause of the problem?
> > > > How to solve the problem?
> > >
> > > DRBD version,
> >
> > 8.3.10
>
> The in-tree or the out-of-tree thing?

In-tree.

We have the options:
disk {
no-disk-barrier;
no-disk-flushes;
no-disk-drain;
}

and
mount-options: type ext4 (rw,noatime,data=writeback)

Should we use disk-barrier? Or the mount option nobarrier?

Thanks for your help.

--
Dr. Michael Schwartzkopff
Guardinistr. 63
81375 München

Tel: (0163) 172 50 98
Fax: (089) 620 304 13
Re: BarrierAck errors on my DRBD [ In reply to ]
> On Thu, Sep 15, 2011 at 11:14:53AM +0200, Michael Schwartzkopff wrote:
> > > On Thu, Sep 15, 2011 at 10:29:47AM +0200, Michael Schwartzkopff wrote:
> > > > Hi,
> > > >
> > > > After fixing an other error I started to see the following error on
> > > > my DRBD:
> > > >
> > > > kernel: block drbd0: BAD! BarrierAck #2778697761 received, expected
> > > > #2778697760!
> > > >
> > > > Can anybody explain what the BarrierAck error means?
> > > > What is the cause of the problem?
> > > > How to solve the problem?
> > >
> > > DRBD version,
> >
> > 8.3.10
>
> The in-tree or the out-of-tree thing?
>
> > > kernel version,
> >
> > 2.6.39.3
> >
> > > what "other error",
> >
> > Since we have a quite limited test setup with virtual machines I saw lots
> > of
> >
> > kernel: block drbd0: PingAck did not arrive in time.
> >
> > I solved that problem
> > - limiting sync to 1M
> > - setting ping-int to 30 (seconds)
> > - setting ping-timeout to 20 (1/10th of seconds, i.e. 2 sec)
> >
> > > what did you do to trigger this,
> >
> > Nothing. It happens about twice per hour.
> >
> > > can you easily reproduce, etc.
> >
> > Yes. That is the annoying problem.
> >
> > > Most likely: Upgrade.
> >
> > From kernel 2.6.39 to?
>
> From 8.3.10 in-kernel (I assume)
> to 8.3.11-3 out-of-tree?

Hi,

The problem was fixed after we recompiled the 8.3.11 module and used this
instead of the in-kernel (2.6.39) module.

Thanks Lars for the help!

Greetings,

--
Dr. Michael Schwartzkopff
Guardinistr. 63
81375 München

Tel: (0163) 172 50 98
Fax: (089) 620 304 13