[ClusterLabs] 答复: Re: 答复: Re: clone resource not get restarted on fail

[Dejan Muhamedagic]

On Wed, Feb 15, 2017 at 09:21:50AM -0600, Ken Gaillot wrote:
> On 02/15/2017 03:57 AM, he.hailong5 at zte.com.cn wrote:
> > I just tried using colocation, it dosen't work.
> > 
> > 
> > I failed the node paas-controller-3, but sdclient_vip didn't get moved:
> 
> The colocation would work, but the problem you're having with router and
> apigateway is preventing it from getting that far. In other words,
> router and apigateway are still running on the node (they have not been
> successfully stopped), so the colocation is still valid.
> 
> I suspect that the return codes from your custom resource agents may be
> the issue. Make sure that your agents conform to these guidelines:
> 
> http://clusterlabs.org/doc/en-US/Pacemaker/1.1-pcs/html-single/Pacemaker_Explained/index.html#ap-ocf

Just to note that there's also "The OCF Resource Agent
Developer’s Guide":

https://github.com/ClusterLabs/resource-agents/blob/master/doc/dev-guides/ra-dev-guide.asc

It should spell all the details for RA developers.

Thanks,

Dejan