We've IBM storwize V7000 Storage where the firmware is going to be upgraded to latest version.. This will require a reboot to the storage.
I've got arround 10 LUN's (almost 10TB) coming from this Storage and mounted on Red Hat Linux 6.4 server.
The server is a part of HA-LVM Cluster running DM Multipathing and it's primary node. I've QLogic QLE2560 HBA card connected on server.
My qustion is what will be the effect on these LUN's when storage goes for reboot.
Shall I ummount all of them and because it's a HA-LVM Cluster Can I only stop the clustered service making use of the LUN's
As I have not done that before I want to be ready with precautions that I can take before sotrage is rebooted and I can mount them again once storage comes online.
Thanks
You should really be talking with your IBM partner on cases like these, as they're the best ones to answer these kind of questions.
That being said - the storwize v7000 is a dual-controller system. As long as you have dual paths (you mentioned DM multipathing) to each controller then a storage processor upgrade should reboot one controller at the time, failing the active LUN's over to the other controller.
I suggest that you read the service manual for your array, which should outline this and give you a procedure for doing upgrades.
The V7000 will upgrade and reboot the nodes one at a time with a 30 minute pause enforced between one half of the nodes in the cluster and their respective partners.
This means that your host will see half of the paths dissapear from the fabric and return, then a after half an hour, the other paths will do the same. As long as your DM Multipathing configuration is correct; the volumes will remain online and the upgrade can be performed concurrently to host operations.
See here for Fibre Channel multipath configuration: http://www-01.ibm.com/support/knowledgecenter/ST3FR7_7.4.0/com.ibm.storwize.v7000.740.doc/svc_linmultipathovr_21bb33.html?cp=ST3FR7%2F0-3-2-1-7-2-0&lang=en
An impact you will see is that while a node is offline for upgrade, the I/O group it is in will run cache disabled. Or if this upgrade is from a release before 7.3.0.0 to 7.3.0.0 or later, then cache will be disabled for the entire upgrade. Some users choose to quiesce host activity during upgrades to mitigate any potential performance impact, you may or may not choose to do so.