Has anyone got single image to work?

[UserName1]

We are working to get single image on a new vSAN cluster and this is our first time working with it rather than baselines.

Has anyone got it to work from start to finish?

Apply ESXi update works - reboots server as expected

Apply HPE customisation, works as expected

Apply HPE SPP firmware. No chance.  Will apply the first 4 for the SPP and not reboot the host, just sits there.  SUT is set to AutoDeploy as detailed in the documentation.

I was hoping this was a one click and away it goes but its just an interactive and needy as the old baseline method.

Thanks.

[tech3d]

Hi  UserName1,

Updating driver + firmware using iSUT is a one click process with the exception that the server reboot is to be done manually. This is recommended to avoid unexpected server reboots which might not be in maintenance mode. Once all the tasks that doesn't need server reboot are completed, the server profile will display the driver and firmware update status (on the server profile) as pending reboot. Since you are performing a fresh deployment, you may very well use 'AutoDeployReboot' mode. You just need to ensure that the server on which the SSP (Synergy service pack) is being applied is in maintenance mode.

Here are some urls that should be useful:

https://www.hpe.com/us/en/servers/hpe-esxi.html

https://support.hpe.com/hpesc/public/docDisplay?docId=a00060084en_us&docLocale=en_US

https://support.hpe.com/hpesc/public/docDisplay?docId=c05212310

(please ignore if you already have these above links)

[UserName1]

Thanks for the reply - unfortunately what you say isn't the case.

Single Image wont let you have the SUT status set as AutoDeployReboot, when it is set to this it will not pass the pre-check scan and error SUT is not set to AutoDeploy.

The servers are in maintenance mode in both OV and VC from the the start of the deployment.

So are you saying that single image is a manual process and I have to reboot the ESXi host, SUT doesnt know when to reboot it or if it is in maintenance mode?

[UserName1]

Also when you see it says 'reboot pending' on the profile do you reboot it via OV of VC?

We did it via OV and it failed the update in VC.

[PriitP]

I had single image updating working until latest SPP. There is an issue in ilo5 firmware 2.55 or newer which behaves exactly like you described. I have a support case open on this and engineering acknowleged it. For workaround you can manually reboot esxi via VC after these 4 firmwares are updated. Monitor the status in OV. After reboot OV is able to finish firmware updates and change status to reboot pending. It then signals this to OV4VC and VC reboots the esxi and task finish successfully.

[UserName1]

That's really interesting.  I thought I did have it working on another cluster I had early last year, that is running iLO 2.42.  One of the first things it updates as part of the 'firmware' stage after ESXi and Firmware and Driver addons is the iLO to 2.55.

Have you tried it on the latest 2.60 version by any chance?

Is it worth me also raising a case to get some more traction on a fix (but opening a OneView ticket is almost impossible)?

[ChrisLynch]

Please do open a support case, so a support engineer can examine the support logs.  Yes, this should work, and simply by specifying the baseline within the profile and ensuring that the installation method is Drivers using Smart Update Tools (which is the online method.)  You do need to make sure that SUT is set to AutoDeployReboot if you want it fully automated, including the reboot of the host.  However, you will need to ensure that the host is put into maintenance mode first.

[T_1_6 · Accepted solution]

I have never seen that autodeployReboot... ours are all set for "Autodeploy" without the reboot part for Mode.

I am sure that is what was written in the HPE docs for vLCM.

Is that an older command perhaps that is the same as simply "autodeploy", or another option entirely?

All our clusters show compliance for vLCM with the OV4VC plug in, in vCenter.

[T_1_6]

My experience of vLCM with the 2020.11 SPP was a diasaster too.
The point of vCLM is full automation as it works its way through the cluster one by one, putting the hosts into maintenance mode and doing all the patching.

Whether there were issues with the Intel 10Gbe NIC firmware and also ILO2.55, I have no idea, all our hosts needed about 6 reboots each to complete. SUT would simply time out and give errors, I did post up the issues I saw if you find my other posts about this. Very curious, hope the next one works as designed, vCLM would be fantastic and a real time and hassle saver!

[UserName1]

Are you using the standard SPP or are you using the VUP version?  We have just been advised to look at using the VUP ones for vLCM as they are designed for VMware only and don't have the non VMware parts in the SPP.

So to answer my original question, yes people have got it to work but the lastest iLO/OV version (6.4) seem to break it.

Lets hope HPE fix the issues in the next release......