Ocfs2 users - Apr 2006 - What happens if Ethernet connection dies

Dear List,

I now have a different problem, which might cause us to stop using
o2cfs.

We still have two HP-machines, redundantly connected to an HP-san with
an EVA 6000. Each has two network cards which are bundled with bonding.o
as active-standby. They are connected to a switch.

Now, if there is a problem with the network, or if one machine is
disconnected from the switch: Both machines are self-fencing!

Installed are:
rhenad:~ # rpm -q ocfs2-tools
ocfs2-tools-1.1.4-0.5
rhenad:~ # rpm -q ocfs2console
ocfs2console-1.1.4-0.5

We cannot really make an update, at least not a kernel update, as we
will loose SAN certification and experiments show, that the SAN and
failover won't work than. Oh yes, all this is from SLES9 SP3

Question: Of course, our main idea to use ocfs was, that if one node
crashes or looses network connection or something, that the other one
survives. At the moment it seems, that if one node has problems, both
are self-fencing.

What is the official opinion here, and what can we do to address that
problem?

Would having each link of the fault-tolerent pairs on a different lan
switch (for redundancy) help? This way, a lan-switch lockup (as it
happened on us) won't be a problem, however, if one(!) machine crashes
or looses network connection, is there a chance for an ocfs2 cluster to
survive?

Thanks in advance,
Michael.

---

Dr. Michael Weller

ITZ Informationstechnologie GmbH
Consulting/Systemengineering
Bismarckstrasse 57
D-45128 Essen

Phone Office  +49 201 24714 28
FAX   Office  +49 201 24714 33
Phone Mobile  +49 172 2178078
E-Mail        mailto:michael.weller at itz-essen.de


-------------- next part --------------
A non-text attachment was scrubbed...
Name: Michael Weller.vcf
Type: text/x-vcard
Size: 570 bytes
Desc: Michael Weller.vcf
Url :
http://oss.oracle.com/pipermail/ocfs2-users/attachments/20060420/c66d3ea4/MichaelWeller.vcf

In a two node situation, if the nodes miss network hbs, but
still can see each other on disk, the higher node will fence itself.

Weller, Michael wrote:
> Dear List,
>
> I now have a different problem, which might cause us to stop using
> o2cfs.
>
> We still have two HP-machines, redundantly connected to an HP-san with
> an EVA 6000. Each has two network cards which are bundled with bonding.o
> as active-standby. They are connected to a switch.
>
> Now, if there is a problem with the network, or if one machine is
> disconnected from the switch: Both machines are self-fencing!
>
> Installed are:
> rhenad:~ # rpm -q ocfs2-tools
> ocfs2-tools-1.1.4-0.5
> rhenad:~ # rpm -q ocfs2console
> ocfs2console-1.1.4-0.5
>
> We cannot really make an update, at least not a kernel update, as we
> will loose SAN certification and experiments show, that the SAN and
> failover won't work than. Oh yes, all this is from SLES9 SP3
>
> Question: Of course, our main idea to use ocfs was, that if one node
> crashes or looses network connection or something, that the other one
> survives. At the moment it seems, that if one node has problems, both
> are self-fencing.
>
> What is the official opinion here, and what can we do to address that
> problem?
>
> Would having each link of the fault-tolerent pairs on a different lan
> switch (for redundancy) help? This way, a lan-switch lockup (as it
> happened on us) won't be a problem, however, if one(!) machine crashes
> or looses network connection, is there a chance for an ocfs2 cluster to
> survive?
>
> Thanks in advance,
> Michael.
>
> ---
>
> Dr. Michael Weller
>
> ITZ Informationstechnologie GmbH
> Consulting/Systemengineering
> Bismarckstrasse 57
> D-45128 Essen
>
> Phone Office  +49 201 24714 28
> FAX   Office  +49 201 24714 33
> Phone Mobile  +49 172 2178078
> E-Mail        mailto:michael.weller at itz-essen.de
>
>
>   
> ------------------------------------------------------------------------
>
> _______________________________________________
> Ocfs2-users mailing list
> Ocfs2-users at oss.oracle.com
> http://oss.oracle.com/mailman/listinfo/ocfs2-users
>

BTW: I wonder why, the fs could easyly mount the drive read only?

Bernd

-----Original Message-----
From: ocfs2-users-bounces at oss.oracle.com
[mailto:ocfs2-users-bounces at oss.oracle.com] On Behalf Of Sunil Mushran
Sent: Friday, April 21, 2006 12:39 AM
To: Weller, Michael
Cc: ocfs2-users at oss.oracle.com
Subject: Re: [Ocfs2-users] What happens if Ethernet connection dies

In a two node situation, if the nodes miss network hbs, but still can
see each other on disk, the higher node will fence itself.

It could. Just that we have to implement that. :)

Eckenfels. Bernd wrote:
> BTW: I wonder why, the fs could easyly mount the drive read only?
>
> Bernd
>
> -----Original Message-----
> From: ocfs2-users-bounces at oss.oracle.com
> [mailto:ocfs2-users-bounces at oss.oracle.com] On Behalf Of Sunil Mushran
> Sent: Friday, April 21, 2006 12:39 AM
> To: Weller, Michael
> Cc: ocfs2-users at oss.oracle.com
> Subject: Re: [Ocfs2-users] What happens if Ethernet connection dies
>
> In a two node situation, if the nodes miss network hbs, but still can
> see each other on disk, the higher node will fence itself.
>