[ClusterLabs] Antw: Is there a way to ignore a single monitoring timeout
klecho at gmail.com
Fri Sep 1 02:48:14 EDT 2017
Have to disagree here.
I have cases, when for an unknown reason a single monitoring request
never returns result.
So having bigger timeouts doesn't resolve this problem.
On 1.09.2017 09:11, Ulrich Windl wrote:
>>>> Klechomir <klecho at gmail.com> schrieb am 31.08.2017 um 17:18 in Nachricht
> <2733498.Wyt05tt8L0 at bobo>:
>> Hi List,
>> I've been fighting with one problem with many different kinds of resources.
>> While the resource is ok, without apparent reason, once in a while there is
>> response to a monitoring request, which leads to a monitoring timeout and
>> resource restart etc
>> Is there any way to ignore one timed out monitoring request and react only
>> two (or more) failed requests in a row?
> I think you are asking the question in the wrong way: You'll have to choose a timeout value that fails in less than 0.x of all cases when the resource is fine, while it fails in (1 - 0.x) cases when the resource has a problem (assuming the monitor just hangs if there is a problem). It's up to you to select the x (via timeout of monitor) that fits your needs.
> In a nutshell: short timeouts may produce errors when there are none, long timeouts may cause extra delays when there is a problem.
> BTW: Having an extra round of monitoring is equivalent to doubling the timeout value; isn't it?
> Users mailing list: Users at clusterlabs.org
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org
More information about the Users