Andrew,<br><br>The updated description looks nice, but could you please remove my <a href="http://fbsdata.com">fbsdata.com</a> domain name from the man page? Also, the "memcached" OCF script was my own creation, and might not be a good example. Maybe one of the other commonly used example resources like an IP address or mysql or something?<br>
<br>If you guys would like to include my memcache my OCF script in the pacemaker distribution just let me know, I'll clean it up for public use and email it.<br><br>Thanks again!<br><br>--Cal<br><br><br><div class="gmail_quote">
On Wed, Oct 24, 2012 at 7:32 PM, Andrew Beekhof <span dir="ltr"><<a href="mailto:andrew@beekhof.net" target="_blank">andrew@beekhof.net</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
How is this?<br>
<br>
...<br>
<br>
-i, --op-inject=value<br>
Generate a failure for the cluster to react to in the simulation<br>
<br>
Value is of the form<br>
${resource}_${task}_${interval}@${node}=${rc}. Eg.<br>
<a href="mailto:memcached_monitor_20000@m1.fbsdata.com">memcached_monitor_20000@m1.fbsdata.com</a>=7<br>
<br>
-F, --op-fail=value<br>
If the specified task occurs during the simulation, have<br>
it fail with return code ${rc}<br>
<br>
Value is of the form<br>
${resource}_${task}_${interval}@${node}=${rc}. Eg.<br>
<a href="mailto:memcached_stop_0@m1.fbsdata.com">memcached_stop_0@m1.fbsdata.com</a>=1<br>
<br>
The transition will normally stop at the failed action,<br>
save the result with --save-output and re-run crm_simulate with<br>
--xml-file<br>
<br>
...<br>
<br>
EXAMPLES<br>
Pretend the recurring memcached monitor failed on node<br>
<a href="http://m1.fbsdata.com" target="_blank">m1.fbsdata.com</a> and, during recovery, that the memcached stop action<br>
did too<br>
<br>
# crm_simulate -LS --op-inject<br>
<a href="mailto:memcached%3A0_monitor_20000@m1.fbsdata.com">memcached:0_monitor_20000@m1.fbsdata.com</a>=7 --op-fail<br>
<a href="mailto:memcached%3A0_stop_0@m1.fbsdata.com">memcached:0_stop_0@m1.fbsdata.com</a>=1 --save-output<br>
/tmp/memcached-test.xml<br>
<br>
Now see what the reaction to the stop failure would be<br>
<br>
# crm_simulate -S --xml-file /tmp/memcached-test.xml<br>
<div class="HOEnZb"><div class="h5"><br>
<br>
<br>
On Thu, Oct 25, 2012 at 8:43 AM, Andrew Beekhof <<a href="mailto:andrew@beekhof.net">andrew@beekhof.net</a>> wrote:<br>
> On Thu, Oct 25, 2012 at 1:37 AM, Cal Heldenbrand <<a href="mailto:cal@fbsdata.com">cal@fbsdata.com</a>> wrote:<br>
>> Thanks Andrew! My first few attempts at playing around with the failure<br>
>> states are working as expected.<br>
>><br>
>> A few follow-ups below:<br>
>><br>
>><br>
>>> --op-fail isn't the command you want though.<br>
>>> From the man page:<br>
>>><br>
>>> -i, --op-inject=value<br>
>>> $rsc_$task_$interval@$node=$rc - Inject the specified<br>
>>> task before running the simulation<br>
>>><br>
>>> -F, --op-fail=value<br>
>>> $rsc_$task_$interval@$node=$rc - Fail the specified task<br>
>>> while running the simulation<br>
>>><br>
>>> Note the difference between the two descriptions: before vs. while.<br>
>>> --op-inject is the one you want. It is mostly useful for pretending a<br>
>>> recurring monitor failed and seeing what the cluster would do about<br>
>>> it.<br>
>>><br>
>>> --op-fail on the other hand, is used for pretending that part of the<br>
>>> recovery process failed.<br>
>><br>
>><br>
>> Your follow up description here is great, and makes more sense. I was<br>
>> reading "Fail the specified task" as literally, "here's my task, fail it and<br>
>> show me the results" I'd suggest to add a little paragraph in the man page<br>
>> to elaborate these points too.<br>
><br>
> Ok, I'll add that today.<br>
><br>
>> Also, can you tell me what all of the return<br>
>> codes are? Do I have to use integers, or do strings like "error" work?<br>
><br>
> Just integers I'm afraid.<br>
> The full list for OCF agents is here:<br>
> <a href="http://www.clusterlabs.org/doc/en-US/Pacemaker/1.1-pcs/html/Pacemaker_Explained/s-ocf-return-codes.html" target="_blank">http://www.clusterlabs.org/doc/en-US/Pacemaker/1.1-pcs/html/Pacemaker_Explained/s-ocf-return-codes.html</a><br>
> LSB return codes are slightly different.<br>
><br>
>> While we're on the subject of documentation / usability, I would also<br>
>> suggest to split out these two features into more parameters. (What would<br>
>> happen if I named my resource with an underscore?) Maybe something like:<br>
>><br>
>> --op-pre-resource=[primitive name]<br>
>> --op-pre-task=[monitor|start|stop]<br>
>> --op-pre-interval=[integer]<br>
>> --op-pre-node=[hostname]<br>
>> --op-pre-rc=[error|timeout|other stuff]<br>
>><br>
>> Then have similar --op-post-* parameters. Or whatever verbs make the most<br>
>> sense in the spirit of Pacemaker vocabulary. (pre/post, before/after,<br>
>> inject/fail, input/output, etc)<br>
><br>
> The reason for not doing that, is that we wanted to be able to inject<br>
> multiple pre/post failures at a time and see the result.<br>
><br>
>> And, examples are always awesome in man<br>
>> pages too.<br>
>><br>
>> Of course, this is all great future version stuff, but that doesn't help all<br>
>> of the RedHat 6 people that will be using pacemaker 1.1 packages for the<br>
>> next ~10 years until RedHat 7 comes out.<br>
><br>
> Don;t worry, the man page updates we just talked about will be in the<br>
> 6.4 packages :)<br>
><br>
>> So I suppose documenting the old<br>
>> code in the online docs is a Good Thing. :-)<br>
>><br>
>> Thanks again!<br>
>><br>
>> --Cal<br>
>><br>
>><br>
>><br>
>> _______________________________________________<br>
>> Pacemaker mailing list: <a href="mailto:Pacemaker@oss.clusterlabs.org">Pacemaker@oss.clusterlabs.org</a><br>
>> <a href="http://oss.clusterlabs.org/mailman/listinfo/pacemaker" target="_blank">http://oss.clusterlabs.org/mailman/listinfo/pacemaker</a><br>
>><br>
>> Project Home: <a href="http://www.clusterlabs.org" target="_blank">http://www.clusterlabs.org</a><br>
>> Getting started: <a href="http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf" target="_blank">http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf</a><br>
>> Bugs: <a href="http://bugs.clusterlabs.org" target="_blank">http://bugs.clusterlabs.org</a><br>
>><br>
<br>
_______________________________________________<br>
Pacemaker mailing list: <a href="mailto:Pacemaker@oss.clusterlabs.org">Pacemaker@oss.clusterlabs.org</a><br>
<a href="http://oss.clusterlabs.org/mailman/listinfo/pacemaker" target="_blank">http://oss.clusterlabs.org/mailman/listinfo/pacemaker</a><br>
<br>
Project Home: <a href="http://www.clusterlabs.org" target="_blank">http://www.clusterlabs.org</a><br>
Getting started: <a href="http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf" target="_blank">http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf</a><br>
Bugs: <a href="http://bugs.clusterlabs.org" target="_blank">http://bugs.clusterlabs.org</a><br>
</div></div></blockquote></div><br>