2015-10-06 02:07 PM
We have some production volumes that contain a couple luns that are snap mirrored from a production netapp to a test netapp (cdot)
We then take the snap mirrored volume thats in test, create a flexclone of it and present that to Vmware to be presented to some VM's and add the luns as RDMs Then we break split the clone and all is well.
Its not a hard process, but somewhat tedious and repetative. Is this something that can be done in WFA?
Here are the steps we do.
Offline existing volume if it exists
Delete existing volume if it exists
Create a flexclone of the mirrored volume
Add inititators to LUN's
Scan ESX host
add RDM's to the virtual machine
Split the clone
I've looked at a few canned workflows but none seem to do what I want. I figured I'd throw it out to the community and see if someone else has done a WF for this. I've played around with creating my own but the clone volume piece seems to be hanging me up.
2015-10-06 02:24 PM - edited 2015-10-06 03:23 PM
Primarily my ability to not know the difference in my own environments
I was building a test WF that points to my test OCUM and wondered why my production WFA server wasnt querying things properly. Ill keep playing around with this WF, hopefully its something simple.
2015-10-08 12:00 PM
I have the old volume removal, the clone volume, and the split volume done in one workflow. Now Im looking to add the mapping of a lun that was created in the clone volume piece and I thin kthis is where it gets tricky.
When the clone volume happens, that volume contains luns. I want to map those specific luns to a certain iGroup. What would be the best way to add an iGroup to a lun that was created in the existing step? I've done it before when creating new luns, but since the Parameters for Map Lun are requiring the LUN definition, Im not sure what to look for. Its asking for the full path, which I can enter, but /vol/ needs to be added to the full lun. Obviously /vol/$servername variable wont work though. See attached screen shot.
2015-10-09 02:57 AM
@ Obviously /vol/$servername variable wont work though. See attached screen shot.
The parameters in Resource Selection are MVEL expressions. You can give like below:
2015-10-09 09:27 AM
The mapping of the lun is where I get a little confused. Since the volume, nor the luns exist prior to a step in my workflow, its hard to add them automatically. I've also tried the search parameter 'filter luns by key', filing-in attributes, and using an existing lun with the '/vol/'+$servername comand and the "/vol/"+$servername and neither seem to work.
This seems to be the same problem I'm having but havent got an answer to my question there.
2015-10-13 12:52 AM
Leve the last thread. We will try to resolve it for you here.
Okay let me get your requirements. You workflow needs to do the following
1. Create a volume
2. Create a lun on that volume
3. Map the created lun on an existing igroup
Is this correct?
2015-10-13 07:44 AM
I've gotten most steps worked out... but having issues with a few still
1. Offline existing volume if it exists
2. Delete existing volume if it exists
3. Create a flexclone of the mirrored volume. Mirrored volume has multiple LUNS
4. Add igroup to all the LUN's
5. Scan ESX host
6. add RDM's to the virtual machine
7. Split the clone
I've gotten 1, 2, 3 and 7 done, the ones that need to be done in VMware arent the big issue. Steps 3, and 4 are the ones that Im having issues with.
2015-10-27 02:40 AM
Okay so you are facing problems in step 4 5 6, right?
Step5 and 6 are totally independent tasks. You need Vmware commands and it should be done.
For step 4 you can use looping on command "Map Lun" with Member In Group. You can have a finder with a filter and the filter to get all luns in a given volume. Now every single lun that is 'found', will have to go through the command Map Lun to map it to the initiator you have given.
If you have your workflow with you, you can post it here. And I can look into it.
If you want a ready-made workflow to meet all your requirements with total support, your company needs to get a NetApp PS engineer.