Releases : z/VM 6.1 Red hat 6.2
We are in the startup process to use SAMP and have no knowledge regarding the product. What we want to achieve is the
We have 2 z/VM systems. In each of z/VM system there is a linux machine that shares the same disks. Only one of these 2 linux machine should be started/logged on in z/VM. When we are doing maintenance we would like to stop the active machine and start it on the other site.
Can anyone give some advice on how to set this up? As earlier mentioned we are novice whith this product.
Pinned topic Question Regarding moving z/VM linux machines.
Answered question This question has been answered.
Unanswered question This question has not been answered yet.
Updated on 2012-11-23T08:30:24Z at 2012-11-23T08:30:24Z by SystemAdmin
Re: Question Regarding moving z/VM linux machines.2012-11-23T08:30:24ZThis is the accepted answer. This is the accepted answer.
- SystemAdmin 110000D4XK
if you use GDPS/XDR to manage VM, I recommend that you use GDPS to control this scenario in order to have GDPS as single point of control.
This means that you would need to set up the two Linux systems as xDR production guest systems. For each Linux system, create a single node cluster (SA MP cluster with only one system), and set up xDR for a production guest as described in the GDPS documentation (GDPS/PPRC V3R9:
Extensions for Heterogeneous Platforms).
This makes GDPS aware of the two Linux systems and you will be able the see these systems in GDPS panels and can control them via GDPS.
Additionally (optionally) you can set up an SA MP policy in order to do a coordinated start/stop of the software stack on the Linux systems.
For unplanned outages of a guest, you are warned via an GDPS operator prompt.
For your planned maintenance scenario, I recommend to write a GDPS script (GDPS PLANNED ACTIONS), which stops the Linux system on the z/VM system and site 1 and starts the orther Linux system on the z/VM system on site 2.
E.g. if you want to stop z/VM guest prod1 on z/VM system VM1, the GDPS script statement looks like this:
CLUSTER='XDR STOP CLUSTER(VM1) NODE(prod1)'
if you want to start z/VM guest prod2 on z/VM system VM2, the GDPS script statement looks like this:
CLUSTER='XDR START CLUSTER(VM2) NODE(prod2)'
Then you can write a second script for "go home" which stops the Linux system on site 2 and starts the other Linux system on site 1 again.
If you need more details, please let me know.