[Pacemaker] Question about OCF RA reload behaviour

Felix Zachlod fz.lists at sis-gmbh.info
Wed Oct 29 06:37:26 UTC 2014


Am 29.10.2014 01:39, schrieb Andrew Beekhof:
>> I just have one question about how a resource agent should behave on "reload" invoked if the resource is currently stopped. Should the resource be started or remain stopped?
>
> I don't think that is defined. Certainly pacemaker wont trigger that case unless the resource fails between the reload command and the monitor operation immediately preceding it.
>
>> I did not find anything about that in the documentation, in gernel there are not mant informations about "reload".
>>
>> I just noticed that my resource remained up after testing it with ocf-tester and this was because I implemented it the way that it would be started if reloaded in stopped state and ocf-tester invokes reload as last action. So I wondered if this was correct?
>
> It seems reasonable to me

Thanks, Andrew, for making this clear.

regards, Felix




More information about the Pacemaker mailing list