Issues regarding commiting the softwares on CX 4

Hello Bonnu,

Creating a LUN is the most basic of any activity.
Could you let us know what error is the array throwing while you create a LUN?

What are the softwares that need to be comitted?

Regards,

Sheron Godfred.

Hi Bonnu,

Commit does not trigger a SP reboot.

It just prevents you from rolling back to the previous FLARE version (flare 28 in your case).

Please see attached file for the error message and here are the software list that needs to be committed.

Snapview

SAN copy

policy engine

Mirriw view

Flare-Operating system

Fast cache configs

EFD Cache engine

DriveFirmware

BASE

RRR is correct, if I remember we need no reboot to commit, however the upgrade ill cause the SP to boot one at a time like 4 times or so.

Regards

Sheron Godfred

Are you sure about this... because i just talked to a support person about this and they told it will reboot the SP's

Well, I think it was twice per SP, so 4 in Total, right?

Yes RRR, if everything goes as planned then its 2*reboots per SP.

And unless you commit the FLARE it aint do any good Bonnu.

U've gotta commit it so that you can get the feature of the latest FLARE.

All the best and keep us posted.

Regards,

Sheron Godfred

I guess i will commit it from the Unisphere.. hopefully nothing should go down. Thanks much for your help and answers.

Having upgraded the FLARE code, you now have two options:

1) Commit the FLARE code (and any layered apps that are flagged: COMMIT_REQUIRED):

- This in itself doesn't require a reboot; however, the upgrade leading up to this point did as already mentioned

- This is the only way you'll be able to make any changes within Unisphere as you've experienced already attempting to create a LUN

- For now you basically have the ability to just view your existing environment but from the perspective of the new Unisphere view having upgraded from FLARE 28 to FLARE 30 (assuming you weren't already managing your previous environment with the Unisphere Client)

2) Do not commit the FLARE code but reboot the SP's

- This is supposed to effectively revert you back to the original version prior to the upgrade (FLARE 28 in your case).  I personally haven't had to do this myself though in the field.

Sheron and RRR,  I think for newer systems SP reboots only once during the upgrade, not sure in case of first generation CX.

During this procedure, each storage processor will reboot independently.  Therefore, it is imperative that all servers have access to both storage processors in the array and have properly configured/working failover software or they will lose access to the array.

Perform a “disruptive” upgrade. If you select this, then both SPs will be rebooted at the same time.  Select this option only if the customer has arranged for all attached hosts to be off-line. 

You will see Engineering Mode displayed.

I think each SP reboots twice, so in all 4 reboots.

Next time when we do an upgrade I'll confirm howmany reboots

Regards,

Sheron Godfred

In our case SP rebooted twice and when we commited the software.. it didnt reboot any of the SP's. I guess for older generation CX it doesnt happen such.

The Storage Processor during a FLARE upgrade only reboots once (Per SP). It is whole different scenario when one needs to reboot any particular Storage Processor, if they encounter any issue during an upgrade.

In total Reboot Per SP equal to 1

Total Reboot: 1

Primary SP: 1

Peer SP: 1

Refer to USM Screenshot(s):

and next screen is a Summary Screen