HPE Aruba 2930F in a stack - Updating Firmware

[Grey-Fox]

Hi all,

This is my first post here. I was hoping someone would be able to help me please. I've had a bit of a Google and even read a few posts on here that are quite similar, but not exactly the same. Unfortunately, I've not been able to find an answer to my question. Most things I've read suggest that it should just work.

A bit of background first:

We have several HPE Aruba 2930F-48G-740W-PoE+-4SFP+ (JL558A) switches on site, and I thought it would be a good idea to update the firmware. It's half term, so I can get away with doing reboots, or scheduling them to run over night. Most are running WC.16.06.0006. However, I've downloaded WC.16.08.0001 (although I've just seen that WC.16.08.0002 is out) and installed that on a few of them to test. All are working fine. The ROM version is WC.16.01.0004. This is true for the ones that have been updated as well.

I downloaded the firmware and manually uploaded this to all the switches via the GUI (we don't currently have Airwave, or any other method that I'm aware of to manage the switches) and scheduled reboots over the next few days. However, I've just found that one of these switches, a VSF stack (two switches), is a bit upset with the update. In the event list, I have a 'Major': 'Reboot of Member ID 2, Wrong OS reboot'.
On the GUI, I see Primary Software is WC.16.08.0001 (Booted), but CLI says WC.16.06.0006. I'm assuming the update rolled itself back/didn't apply?

The update has worked fine on the single switches. Looking at another one of my stacks, this time three switches, the firmware update has been successful too. However, this is running WC.16.01.0005 ROM. In addition, I manually provisioned the switches in this stack. This was done a few weeks ago, and were purchased separately for a new building.

My question is, how do I update the firmware on VSF stacked switches please? Am I doing something wrong, or could this just be an issue with this one stack? Being in a stack, I assumed the firmware would apply to all switches. I think this was the case for our previous switches - non-HP.

Thank you in advance.

Kind regards

Grey Fox

[parnassus · Accepted solution]

First of all I strongly suggest you to work using CLI (Telnet or SSH, whatever you prefer)...once in CLI on the VSF IP Address perform a show vsf detail and a show flash to exactly understand the status of your Aruba 2930F based VSF from the VSF and Flash standpoints.

Once you have a clear view it's a matter of:

• Verify with show flash the Primary and Secondary Flash software versions of your VSF.
• Upload to VSF the new ROM image (WC.16.08.0002) on the Primary or Secondary Flash area (as example - on Primary - through the CLI command: copy tftp flash ip-address-of-tftp-server WC_16_08_0002.swi primary).
• Verify with show flash that the Primary (or Secondary) Flash are is correctly filled with the new ROM image (note that the uploaded ROM code is automatically copied to each of the other VSF members).
• Personally here I will wait at least two minutes.
• Execute the CLI Command boot system flash primary (if the upload of the new ROM happened to the Primary Flash area) <- this command will reboot your VSF into primary where you've copied the new ROM image.

Read this guide

[Grey-Fox]

Hi Parnassus,

Thank you for your reply.

I've used, 'show vsf detail' and 'show flash' previously, which is how I found the differing software versions. Although, the GUI seems to reflect whatever firmware version you've updated to, despite not having restarted., instead of the live version. As it happens, I think updating the firmware via the GUI again has kicked this stack into gear. It's no longer showing errors after rebooting and the firmware version is up to WC_16_08_0001, as expected.

Thank you for your advice though. I have set up a TFTP server and tested updating a switch, the way you suggested. So, I'll use SSH to update them in the future.
I'll update the firmware and schedule a reboot for Sunday night, on a couple of my other stacked switches. Hopefully these will update correctly and I can put this down to a one-off issue.

Thanks

[Grey-Fox]

Hi,

I was in the middle of posting an update on Monday, as it didn't look like this had solved my issue. All the stacks that were rebooted over the weekend were showing the same error:

Major - System - ST1-STBY - 'Reboot of Member ID 1, Wrong OS reboot'.

Under 'show system', 'show version' and 'show vsf detail':

Software revision : WC.16.06.0006

Under 'show flash':

Primary Image : 29616419 11/21/18 WC.16.08.0001
Secondary Image : 28793152 12/22/17 WC.16.05.0004

So, the updated file was there, but not being taken by the stack. Then I realised I was still using 'reboot', instead of 'boot system flash primary', as you'd suggested. The uptime on the second member didn't match that of the first, which was being rebooted with the 'reboot' command. The second was just being promoted to first member, and not restarting. The 'boot system flash primary' command reboots the whole stack, which then allows both of them to be updated.

I have a couple of 'CMDR 2's, so have scheduled another reboot for those. But otherwise, everything is looking good. Thank you again for your help.

[parnassus]

Glad you solved! yes...boot system flash is not like doing just a reboot.