<html><body><p><font size="2">Hi Andrei,<br><br>What if we had no hardware device and all we wanted to do was run a single command when fencing off a failed node. Would that be possible?<br><br>Are fencing operations dependent on underlying hardware for Pacemaker?<br><br>Thanks,</font><br><br><b><font size="2">Phil Stedman</font></b><br><font size="2">Db2 High Availability Development and Support<br>Email: pmstedma@us.ibm.com</font><br><br><img width="16" height="16" src="cid:1__=8FBB0F29DFFC8A4E8f9e8a93df938690918c8FB@" border="0" alt="Inactive hide details for Andrei Borzenkov ---07/30/2020 11:02:58 PM---30.07.2020 22:17, Philippe M Stedman пишет: >"><font size="2" color="#424282">Andrei Borzenkov ---07/30/2020 11:02:58 PM---30.07.2020 22:17, Philippe M Stedman пишет: ></font><br><br><font size="2" color="#5F5F5F">From: </font><font size="2">Andrei Borzenkov <arvidjaar@gmail.com></font><br><font size="2" color="#5F5F5F">To: </font><font size="2">developers@clusterlabs.org, Philippe M Stedman <pmstedma@us.ibm.com>, nwahl@redhat.com</font><br><font size="2" color="#5F5F5F">Cc: </font><font size="2">Toby Haynes <thaynes@ca.ibm.com>, Gerry R Sommerville <gerry@ca.ibm.com>, Alan Y Lee <ykalee@ca.ibm.com></font><br><font size="2" color="#5F5F5F">Date: </font><font size="2">07/30/2020 11:02 PM</font><br><font size="2" color="#5F5F5F">Subject: </font><font size="2">[EXTERNAL] Re: [ClusterLabs Developers] How to implement fencing agent with no associated hardware device with Pacemaker?</font><br><hr width="100%" size="2" align="left" noshade style="color:#8091A5; "><br><br><br><tt><font size="2">30.07.2020 22:17, Philippe M Stedman пишет:<br>> <br>> Thanks Gerry.<br>> <br>> Hi Reid,<br>> <br>> The shared storage solution we are using has clustering capabilities of its<br>> own built into it and is able to remotely fence off the lost node, all we<br>> need to do is run the command to expel/fence the lost node as part of our<br>> own custom fencing agent on the surviving node.<br>> <br>> FYI, the shared storage solution I am referring to here is IBM Spectrum<br>> Scale.<br>> <br><br>...<br><br>> <br>> On Wed, Jul 29, 2020 at 8:34 PM Philippe M Stedman <pmstedma@us.ibm.com><br>> wrote:<br>> Hi ClusterLabs developers,<br>> <br>> I am looking into how to develop a fencing agent for Pacemaker that is<br>> not associated to any underlying hardware device. In our case we have<br>> two servers (we will expand to more in the future) which have access to<br>> shared storage. When one of the two nodes fails, we expect the<br>> surviving node to invoke our user-defined fencing agent and run a<br>> series of commands which will "expel" the lost host from accessing<br>> shared storage.<br>> <br>> Do you have any advice on how to go about implementing such a solution?<br>> All the examples I can find online revolve around using some sort of<br>> underlying hardware device to implement fencing.<br>> <br><br>In this case your fencing *is* associated with specific underlying<br>hardware device. Just start with any of existing fencing or stonith<br>agents and fill in commands to your underlying hardware to fence off<br>node. You probably will need some sort of mapping from node name to<br>whatever is required by your hardware to identify it.<br><br></font></tt><br><br><BR>
</body></html>