On Mon, May 10, 2010 at 12:09 PM, Gianluca Cecchi
<[email protected]> wrote:
> Hello,
> suppose I have group of resources named G1 and a resource named R2.
> I define an order R2 after G1 and a colocation constraint of -inf so
> that they run on different nodes (2 nodes overall).
> At runtime I have G1 on node1 and R2 on node2, correctly.
> I do a move of G1, so that I would have at the end G1 on node2 and R2 on 
> node1.
>
> But suppose node1 doesn' t satisfy requirements for running R2 (for
> example a file system without the mount point defined on node1).
>
> Is it expected behaviour that I get:
> - R2 stops correctly on node2
> - G1 starts correctly on node2
> - R2 then, based on constraint and order, tries to start on node1 but fails

So far, yes

> - G1 too goes in stopped status, so that I now have both G1 and R2 stopped

If I understood your description correctly, then this last part is not expected.
Please create a bug and attach a hb_report archive for the interval
covered by the test case.

> Other questions:
> - If instead I had suddenly run a move and then unmove of G1, would it
> have tried to run on node1 again or is it the failure itself of R2 to
> cause the stop of G1 too?

It depends if you had any ordering constraints in the mix also.
But generally I'd have expected G1 to remain active somewhere.

> - Is there a way to tell that even if R2 fails, then don't impact on G1?

Without an ordering constraint it should do just that.

> - Is there a way to run a move command that automatically does an
> unmove (what normally is called a relocation of a service, which in
> general doesn't imply to mark the node as unable to get the resource
> again back).

Not unless the shell does something like this.
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to