<p dir="ltr">This is true for opt-out clusters. Opt-in cluster only rely on location constraints.</p>
<p dir="ltr">Regards</p>
<div class="gmail_quote">Le 4 avr. 2016 08:46, "Ulrich Windl" <<a href="mailto:Ulrich.Windl@rz.uni-regensburg.de">Ulrich.Windl@rz.uni-regensburg.de</a>> a écrit :<br type="attribution"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi!<br>
<br>
Actually form my SLES11 SP[1-4] experience, the cluster always distributes<br>
resources across all available nodes, and only if don't want that, I'll have to<br>
add constraints. I wonder why that does not seem to work for you.<br>
<br>
Regards,<br>
Ulrich<br>
<br>
>>> Ferenc Wágner <<a href="mailto:wferi@niif.hu">wferi@niif.hu</a>> schrieb am 02.04.2016 um 10:28 in Nachricht<br>
<<a href="mailto:87egao9zyc.fsf@lant.ki.iif.hu">87egao9zyc.fsf@lant.ki.iif.hu</a>>:<br>
> Ken Gaillot <<a href="mailto:kgaillot@redhat.com">kgaillot@redhat.com</a>> writes:<br>
><br>
>> On 03/30/2016 08:37 PM, Ferenc Wágner wrote:<br>
>><br>
>>> I've got a couple of resources (A, B, C, D, ... more than cluster nodes)<br>
>>> that I want to spread out to different nodes as much as possible. They<br>
>>> are all the same, there's no distinguished one amongst them. I tried<br>
>>><br>
>>> <rsc_colocation id="cl-test-spread" score="-50"><br>
>>> <resource_set id="cl-test-spread-set" sequential="false"><br>
>>> <resource_ref id="A"/><br>
>>> <resource_ref id="B"/><br>
>>> <resource_ref id="C"/><br>
>>> <resource_ref id="D"/><br>
>>> </resource_set><br>
>>> <resource_set id-ref="cl-test-spread-set"/><br>
>>> </rsc_colocation><br>
>>><br>
>>> But crm_simulate did not finish with the above in the CIB.<br>
>>> What's a good way to get this working?<br>
>><br>
>> Per the docs, "A colocated set with sequential=false makes sense only if<br>
>> there is another set in the constraint. Otherwise, the constraint has no<br>
>> effect." Using sequential=false would allow another set to depend on all<br>
>> these resources, without them depending on each other.<br>
><br>
> That was the very idea behind the above colocation constraint: it<br>
> contains the same group twice. Yeah, it's somewhat contrived, but I had<br>
> no other idea with any chance of success. And this one failed as well.<br>
><br>
>> I haven't actually tried resource sets with negative scores, so I'm not<br>
>> sure what happens there. With sequential=true, I'd guess that each<br>
>> resource would avoid the resource listed before it, but not necessarily<br>
>> any of the others.<br>
><br>
> Probably, but that isn't what I'm after.<br>
><br>
>> By default, pacemaker does spread things out as evenly as possible, so I<br>
>> don't think anything special is needed.<br>
><br>
> Yes, but only on the scale of all resources. And I've also got a<br>
> hundred independent ones, which wash out this global spreading effect if<br>
> you consider only a select handful.<br>
><br>
>> If you want more control over the assignment, you can look into<br>
>> placement strategies:<br>
><br>
> We use balanced placement to account for the different memory<br>
> requirements of the various resources globally. It would be possible to<br>
> introduce a new, artifical utilization "dimension" for each resource<br>
> group we want to spread independently, but this doesn't sound very<br>
> compelling. For sets of two resources, a simple negative colocation<br>
> constraint works very well; it'd be a pity if it wasn't possible to<br>
> extend this concept to larger sets.<br>
> --<br>
> Thanks,<br>
> Feri<br>
><br>
> _______________________________________________<br>
> Users mailing list: <a href="mailto:Users@clusterlabs.org">Users@clusterlabs.org</a><br>
> <a href="http://clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://clusterlabs.org/mailman/listinfo/users</a><br>
><br>
> Project Home: <a href="http://www.clusterlabs.org" rel="noreferrer" target="_blank">http://www.clusterlabs.org</a><br>
> Getting started: <a href="http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf" rel="noreferrer" target="_blank">http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf</a><br>
> Bugs: <a href="http://bugs.clusterlabs.org" rel="noreferrer" target="_blank">http://bugs.clusterlabs.org</a><br>
<br>
<br>
<br>
<br>
_______________________________________________<br>
Users mailing list: <a href="mailto:Users@clusterlabs.org">Users@clusterlabs.org</a><br>
<a href="http://clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://clusterlabs.org/mailman/listinfo/users</a><br>
<br>
Project Home: <a href="http://www.clusterlabs.org" rel="noreferrer" target="_blank">http://www.clusterlabs.org</a><br>
Getting started: <a href="http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf" rel="noreferrer" target="_blank">http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf</a><br>
Bugs: <a href="http://bugs.clusterlabs.org" rel="noreferrer" target="_blank">http://bugs.clusterlabs.org</a><br>
</blockquote></div>