Hi Nils,

Thanks for your prompt reply.

> If OHAC/Sun Cluster prevented access to shared storage from all but one
> node, it would not be possible to use cluster filesystems (such as QFS or
> VxFS CFS) or parallel databases like Oracle RAC.

if Vxfs' cfs feature is used then it would be a scalable cluster and
the shared mount option would be used, you are perfectly right that in
the case of scalable cluster all the node should have access to the
disk. But even in the case VxFS cfs, the other node is not allowed to
mount if the first node has not mounted the filesystem with share
mount option, so I was just wondering if ufs has any such mechanism.

> If the application needs "raw" device access, then, yes, it should
> cooperate. If it doesn't or when in doubt, you can use the OHAC/Sun Cluster
> framework to guarantee that the application will only ever run on one node
> at a time (for instance using Failover_mode HARD as a last resort, see
> r_properties(5)).

as far as I could understand about the different Failover_mode
options, it tells what is behavior if any of the methods (prestart,
start, stop or poststop)  fail. It does say any thing about the device
access in fail-over cluster.

The reason I ask this question is to confirm that in general offline
devices are actually accessible or its just because some of my
configuration issue.
Looks like its a expected behavior that offline devices are accessible
same as online devices in even fail-over cluster.

Thanks again,
ajit



On Tue, Oct 27, 2009 at 1:32 PM, Nils Goroll <slink at schokola.de> wrote:
> Hi Ajit,
>
>> Now, my question is why shared devices is not protected from secondary
>> while is mounted on primary node?
>> if this is expected behavior that means sun cluster entirely relies on
>> application of behave well and if someone by mistake do a dd(1m) on
>> the device from the secondary then everything is gone for the good, is
>> that acceptable behavior?
>
> If OHAC/Sun Cluster prevented access to shared storage from all but one
> node, it would not be possible to use cluster filesystems (such as QFS or
> VxFS CFS) or parallel databases like Oracle RAC.
>
> One could argue, though, that a feature to restrict access (in a
> configurable way) would prevent some administrator errors, but personally I
> would not value such an option much because it would introduce another
> source of error.
>
> I don't quite understand your argument that the protection of shared storage
> "entirely relie(d) on (the) application". I don't think that applications
> should be given permissions to access device nodes if a file system is used.
>
> As long as you use a fail system, the framework will ensure consistent
> access to it either through PxFS (global mount option) or by using "HA
> local" storage with the SUNW.HAStoragePlus resource type.
>
> If the application needs "raw" device access, then, yes, it should
> cooperate. If it doesn't or when in doubt, you can use the OHAC/Sun Cluster
> framework to guarantee that the application will only ever run on one node
> at a time (for instance using Failover_mode HARD as a last resort, see
> r_properties(5)).
>
> Nils
>

Reply via email to