Disk firmware update

Benoit PICARD wrote:

Hello,
We just received some alerts from our PRISM about the disk firmware. Following the alerts we have to upgrade them.

Info : SATA DOM on host XXXXXX has old firmware version: SATADOM-SL 3IE3 firmware version is S560301N. Need to upgrade to the latest firmware version.

We know that the upgrade do an upgrade one disk at a time but i've some doubts about the process to use before to click to upgrade.

https://portal.nutanix.com/#/page/docs/details?targetId=Web_Console_Guide-Prism_v4_7:wc_cluster_firmware_upgrade_wc_t.html

My question is, should we move our VM from the host before to upgrade the disks? should we add our hypervisor in maintenance mode in order to upgrade the disks?
Or the disk firmware update has no impact on VM hosted ?

Thanks in advance all!!!

So it really depends. If you're using AHV, it will evacuate the host before any upgrades occur, you just need to make sure that you've removed any affinity rules for hosts or at least make sure the rules aren't pinning a VM to one host.

If you're using ESXi and don't have DRS, then yes, you'll want to manually evacuate the host and perform a manual upgrade.

I would double check with support about the firmware version. We received a similar alert from NCC on a large number of hosts, and it ended up being a false positive for SATADOM-SL 3IE3. This was resolved with a new version of NCC. If there is a new version available then you can leverage LCM to upgrade.

Hope that helps some.

For more information around LCM you can reference the following:

https://portal.nutanix.com/#/page/docs/details?targetId=Life-Cycle-Manager-Guide-v20:Life-Cycle-Manager-Guide-v20


or watch this
https://youtu.be/CftB7LhStnQ

Hello,

Thank you for your feedback and explanation.

We received these alerts following an upgrade of NCC Check so i supposed we are already in the latest version.

We are not in AHV but in vSphere with DRS Rules.

Regards,

This is a SATADOM firmware upgrade. The reason you got the alert after upgrading the NCC is because in latest NCC we have introduced the check, to check the SATADOM firmware as well.

The SATADOM contains the Hypervisor, in your case ESXi and can be upgraded through LCM framework. LCM runs pre-checks to verify if DRS is enabled and some additional checks as well before starting the upgrade.

You can leverage LCM and it will upgrade the SATADOM firmware one by one on all nodes.

We're seeing a similar alert on Dell XC blocks with R730xd hosts running AHV, appears to be a false positive:



Is there any way to mute these specific alerts? Didn't see it under Alert Policies.

As @nishantdhingra1 reiterated with LCM, this will evacuate a host for you by obeying DRS. If you do not have DRS or running AHV, you must upgrade a node manually. In that case, yes you would need to evacuate your host before beginning the SATADOM upgrade.

If you're put in that position, reference this guide.

https://portal.nutanix.com/#/page/docs/details?targetId=Hardware-Admin-Ref-AOS-v510:bre-manual-firmware-update-c.html


If anyone has answered your question, please remember to mark their response as an answer, so that others can be assisted by your question in the future.

I had a bad history with LCM a few months before where it hanged at the Phoenix and i have to ask for support from Nutanix.
I did LCM inventory after i upgraded to latest AOS and NCC a few days ago and it found a firmware upgrade for my disks.
I tried the upgrade only one server using LCM and it went fine, then i did it for all servers and it went totally fine without any issues.

i think LCM has been a lot better recently and i wish if they could bring all software/firmware upgrades under it.

LCM is definitely the route Nutanix is headed. I think they're wanting to move everything under that umbrella and make it centralized. We've had some issues in the past with BIOS and BMC updates, but we caught it early in the lab. After the spectre meltdown updates came out, the phoenix software didn't account for a second reboot applied after the update and it hosed the upgrade. So we had the manually finish the cluster.

They've come a long way and as a consumer that's only good news! Glad things went well for you!

Hi,
Does anyone know if there is a way to stage SATADOM or other firmwares so we can control the reboots ourselves. So that the updates are sitting on the server wiating for reboot?