Mailing List Archive: Linux-HA: Users

ocf:heartbeat:exportfs fails tests

Index | Next | Previous | View Threaded

external.Martin.Konold at de

Jul 12, 2012, 12:38 AM

Post #1 of 6 (331 views)
Permalink

ocf:heartbeat:exportfs fails tests

Hi,

I am wondering why migrating a simple nfs resource leads to fencing of the cluster node.

Running resource-agents-3.9.2-0.25.5 on SLES11-SP2 fully updated.

ocf-tester -n test -o fsid="1" -o directory="/SHARED/nfs/home" -o options="rw,mountpoint" -o clientspec="10.0.0.0/255.0.0.0" -o wait_for_leasetime_on_stop="true" -o unlock_on_stop"true" /usr/lib/ocf/resource.d/heartbeat/exportfs
Beginning tests for /usr/lib/ocf/resource.d/heartbeat/exportfs...
* rc=0: Monitoring a stopped resource should return 7
* rc=0: The initial probe for a stopped resource should return 7 or 5 even if all binaries are missing
* Your agent does not support the notify action (optional)
* Your agent does not support the demote action (optional)
* Your agent does not support the promote action (optional)
* Your agent does not support master/slave (optional)
* rc=0: Monitoring a stopped resource should return 7
* rc=0: Monitoring a stopped resource should return 7
* rc=0: Monitoring a stopped resource should return 7
* Your agent does not support the reload action (optional)
Tests failed: /usr/lib/ocf/resource.d/heartbeat/exportfs failed 5 tests

Is there anything wrong with my testing?

Is is normal that released resources fail their included ocf-tester tests? Is this problem SuSE-specific?

Best regards

Martin Konold

Robert Bosch GmbH
Automotive Electronics
Postfach 13 42
72703 Reutlingen
GERMANY
www.bosch.com<http://www.bosch.com>

Tel. +49 7121 35 3322

Sitz: Stuttgart, Registergericht: Amtsgericht Stuttgart, HRB 14000;
Aufsichtsratsvorsitzender: Franz Fehrenbach; Geschäftsführung: Volkmar Denner, Siegfried Dais;
Stefan Asenkerschbaumer, Bernd Bohr, Rudolf Colm, Dirk Hoheisel, Christoph Kübel, Uwe Raschke,
Wolf-Henning Scheider, Werner Struth, Peter Tyroller

_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

external.Martin.Konold at de

Jul 12, 2012, 1:07 AM

Post #2 of 6 (321 views)
Permalink

Re: ocf:heartbeat:exportfs fails tests [In reply to]

Hi,

> I am wondering why migrating a simple nfs resource leads to fencing of the cluster node.

Still trying to resolve this.

> * Your agent does not support the reload action (optional) Tests failed: /usr/lib/ocf/resource.d/heartbeat/exportfs failed 5 tests

> Is there anything wrong with my testing?

I figured out that actually putting the cluster in maintenance mode before doing the test helps significantly ;-)

Sorry for the noise!

Best regards

Martin Konold

Robert Bosch GmbH
Automotive Electronics
Postfach 13 42
72703 Reutlingen
GERMANY
www.bosch.com<http://www.bosch.com>

Tel. +49 7121 35 3322

Sitz: Stuttgart, Registergericht: Amtsgericht Stuttgart, HRB 14000;
Aufsichtsratsvorsitzender: Franz Fehrenbach; Geschäftsführung: Volkmar Denner, Siegfried Dais; Stefan Asenkerschbaumer, Bernd Bohr, Rudolf Colm, Dirk Hoheisel, Christoph Kübel, Uwe Raschke, Wolf-Henning Scheider, Werner Struth, Peter Tyroller

_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

external.Martin.Konold at de

Jul 12, 2012, 3:10 AM

Post #3 of 6 (317 views)
Permalink

Re: ocf:heartbeat:exportfs fails tests [In reply to]

-----Original Message-----
From: linux-ha-bounces [at] lists [mailto:linux-ha-bounces [at] lists] On Behalf Of EXTERNAL Konold Martin (erfrakon, RtP2/TEF72)
Sent: Donnerstag, 12. Juli 2012 10:08
To: 'General Linux-HA mailing list'
Subject: Re: [Linux-HA] ocf:heartbeat:exportfs fails tests

Hi,

> > I am wondering why migrating a simple nfs resource leads to fencing of the cluster node.

> Still trying to resolve this.

For readers oft the archive: This can be caused by insufficient timeouts.

ocf:heartbeat:exportfs needs at least 100s timeout for stopping in order to work properly

Best regards

Martin Konold

Robert Bosch GmbH
Automotive Electronics
Postfach 13 42
72703 Reutlingen
GERMANY
www.bosch.com<http://www.bosch.com>

Tel. +49 7121 35 3322

Sitz: Stuttgart, Registergericht: Amtsgericht Stuttgart, HRB 14000;
Aufsichtsratsvorsitzender: Franz Fehrenbach; Geschäftsführung: Volkmar Denner, Siegfried Dais; Stefan Asenkerschbaumer, Bernd Bohr, Rudolf Colm, Dirk Hoheisel, Christoph Kübel, Uwe Raschke, Wolf-Henning Scheider, Werner Struth, Peter Tyroller
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

dejanmm at fastmail

Jul 12, 2012, 8:07 AM

Post #4 of 6 (318 views)
Permalink

Re: ocf:heartbeat:exportfs fails tests [In reply to]

Hi,

On Thu, Jul 12, 2012 at 12:10:28PM +0200, EXTERNAL Konold Martin (erfrakon, RtP2/TEF72) wrote:
>
> -----Original Message-----
> From: linux-ha-bounces [at] lists [mailto:linux-ha-bounces [at] lists] On Behalf Of EXTERNAL Konold Martin (erfrakon, RtP2/TEF72)
> Sent: Donnerstag, 12. Juli 2012 10:08
> To: 'General Linux-HA mailing list'
> Subject: Re: [Linux-HA] ocf:heartbeat:exportfs fails tests
>
> Hi,
>
> > > I am wondering why migrating a simple nfs resource leads to fencing of the cluster node.
>
> > Still trying to resolve this.
>
> For readers oft the archive: This can be caused by insufficient timeouts.
>
> ocf:heartbeat:exportfs needs at least 100s timeout for stopping in order to work properly

stop timeout is advertised as 10s for exportfs. That's probably
too short. But, do you know why does it take so long to stop?
100s sounds excessive.

Thanks,

Dejan

> Best regards
>
> Martin Konold
>
> Robert Bosch GmbH
> Automotive Electronics
> Postfach 13 42
> 72703 Reutlingen
> GERMANY
> www.bosch.com<http://www.bosch.com>
>
> Tel. +49 7121 35 3322
>
> Sitz: Stuttgart, Registergericht: Amtsgericht Stuttgart, HRB 14000;
> Aufsichtsratsvorsitzender: Franz Fehrenbach; Geschäftsführung: Volkmar Denner, Siegfried Dais; Stefan Asenkerschbaumer, Bernd Bohr, Rudolf Colm, Dirk Hoheisel, Christoph Kübel, Uwe Raschke, Wolf-Henning Scheider, Werner Struth, Peter Tyroller
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

cermakm at gmail

Jul 12, 2012, 3:38 PM

Post #5 of 6 (314 views)
Permalink

Re: ocf:heartbeat:exportfs fails tests [In reply to]

On 13 July 2012 01:07, Dejan Muhamedagic <dejanmm [at] fastmail> wrote:
>
> Hi,
>
> On Thu, Jul 12, 2012 at 12:10:28PM +0200, EXTERNAL Konold Martin
> (erfrakon, RtP2/TEF72) wrote:
> >
> > -----Original Message-----
> > From: linux-ha-bounces [at] lists
> > [mailto:linux-ha-bounces [at] lists] On Behalf Of EXTERNAL Konold
> > Martin (erfrakon, RtP2/TEF72)
> > Sent: Donnerstag, 12. Juli 2012 10:08
> > To: 'General Linux-HA mailing list'
> > Subject: Re: [Linux-HA] ocf:heartbeat:exportfs fails tests
> >
> > Hi,
> >
> > > > I am wondering why migrating a simple nfs resource leads to fencing
> > > > of the cluster node.
> >
> > > Still trying to resolve this.
> >
> > For readers oft the archive: This can be caused by insufficient
> > timeouts.
> >
> > ocf:heartbeat:exportfs needs at least 100s timeout for stopping in order
> > to work properly
>
> stop timeout is advertised as 10s for exportfs. That's probably
> too short. But, do you know why does it take so long to stop?
> 100s sounds excessive.
>

I believe 100s will be needed only if you have the
"wait_for_leasetime_on_stop" parameter set to "true".
It's needed only if you use NFSv4. It causes the RA to wait NFS grace
period, which is usually 90 seconds, when stopping.

If you don't use NFSv4, they set this parameter to false and you will
save NFS grace period amount of time.
If you do use NFSv4, you can try configure your NFS server to use
lower grace period - see chapter "6.1. NFSv4 lease time" of this very
good document:
www.linbit.com/fileadmin/tech-guides/ha-nfs.pdf

Kind regards,
Marji

> Thanks,
>
> Dejan
>
> > Best regards
> >
> > Martin Konold
> >
> > Robert Bosch GmbH
> > Automotive Electronics
> > Postfach 13 42
> > 72703 Reutlingen
> > GERMANY
> > www.bosch.com<http://www.bosch.com>
> >
> > Tel. +49 7121 35 3322
> >
> > Sitz: Stuttgart, Registergericht: Amtsgericht Stuttgart, HRB 14000;
> > Aufsichtsratsvorsitzender: Franz Fehrenbach; Geschäftsführung: Volkmar
> > Denner, Siegfried Dais; Stefan Asenkerschbaumer, Bernd Bohr, Rudolf Colm,
> > Dirk Hoheisel, Christoph Kübel, Uwe Raschke, Wolf-Henning Scheider, Werner
> > Struth, Peter Tyroller
> > _______________________________________________
> > Linux-HA mailing list
> > Linux-HA [at] lists
> > http://lists.linux-ha.org/mailman/listinfo/linux-ha
> > See also: http://linux-ha.org/ReportingProblems
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

external.Martin.Konold at de

Jul 19, 2012, 9:50 AM

Post #6 of 6 (310 views)
Permalink

Re: ocf:heartbeat:exportfs fails tests [In reply to]

Hi,

> > For readers oft the archive: This can be caused by insufficient timeouts.
> >
> > ocf:heartbeat:exportfs needs at least 100s timeout for stopping in
> > order to work properly

> stop timeout is advertised as 10s for exportfs. That's probably too short. But, do you know why does it take so long to stop?
> 100s sounds excessive.

FYR: The delay was caused by the option "wait_for_leasetime_on_stop" for ocf:heartbeat:exportfs.

Mit freundlichen Grüßen / Best regards

Martin Konold

Robert Bosch GmbH
Automotive Electronics
Postfach 13 42
72703 Reutlingen
GERMANY
www.bosch.com

Tel. +49 7121 35 3322

Sitz: Stuttgart, Registergericht: Amtsgericht Stuttgart, HRB 14000;
Aufsichtsratsvorsitzender: Franz Fehrenbach; Geschäftsführung: Volkmar Denner, Siegfried Dais;
Stefan Asenkerschbaumer, Bernd Bohr, Rudolf Colm, Dirk Hoheisel, Christoph Kübel, Uwe Raschke,
Wolf-Henning Scheider, Werner Struth, Peter Tyroller

Thanks,

Dejan

> Best regards
>
> Martin Konold
>
> Robert Bosch GmbH
> Automotive Electronics
> Postfach 13 42
> 72703 Reutlingen
> GERMANY
> www.bosch.com<http://www.bosch.com>
>
> Tel. +49 7121 35 3322
>
> Sitz: Stuttgart, Registergericht: Amtsgericht Stuttgart, HRB 14000;
> Aufsichtsratsvorsitzender: Franz Fehrenbach; Geschäftsführung: Volkmar
> Denner, Siegfried Dais; Stefan Asenkerschbaumer, Bernd Bohr, Rudolf
> Colm, Dirk Hoheisel, Christoph Kübel, Uwe Raschke, Wolf-Henning
> Scheider, Werner Struth, Peter Tyroller
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Index | Next | Previous | View Threaded

Interested in having your list archived? Contact Gossamer Threads