Nimble Connection Manager and VMware Lifecycle image

[joebrug52]

I recently upgraded our hosts to ESXi 8.x using Lifecycle Manager. Based on the documentation (which seems a bit inconsistent), I removed the Nimble Connection Manager prior to the upgrade.

After successfully upgrading all hosts to 8.x, I edited the cluster image to include the Nimble Connection Manager component (added via URL), and then attempted to remediate a host. The remediation task reported a failure, but it appears the Nimble component was actually installed.

I ended up having to manually reboot the hosts, and after that, a compliance check showed everything as compliant.

Did I miss something in the process? I was expecting the host to automatically enter maintenance mode, install the Nimble/HPE component, and reboot as part of remediation.

[controles]

Hello,

Your steps are correct. However, in many cases vLCM does not complete the reboot automatically, and it may report an error even though the component has been installed successfully.

On the host, you can often see a pending reboot status, but the system does not restart by itself.
For this reason, a manual reboot is required to complete the process and bring the host back into a compliant state.

Best regards,
Carlos

[buzzsubash]

Hi @joebrug52 
I did check https://support.hpe.com/hpesc/public/docDisplay?docId=sd00004102en_us&page=GUID-823100E4-10E9-4BE8-AD1C-3FBE21CD0656.html and indeed it is bit confusing.

But can you please tell me what was the original ESXi version and connection manager prior upgrading ESXi? Once I have this info, will create an internal ticket with documentation team to see if some additional details can be added. 

Alternatively, if you would like to investigate why remediation task failed and to make sure NCM installation was completed successfully, kindly create a support ticket, we may need support bundle from vcenter/esxi host etc.

[controles]

Hello,

I agree with that response. From experience, confirming the original ESXi version and identifying when the Connection Manager upgrade was performed really helps to put the issue into context.

If the collection task failed, opening a support case makes sense to verify whether NCM was installed correctly or if something was left incomplete. In many cases, reviewing the internal logs together with HPE Support is the fastest way to move forward.

Thank you for the clarification.

Regards,
Carlos

[joebrug52]

I initially upgraded the Nimble Connection Manager from 7.0.2 to 7.0.3 while running ESXi 7.0 Update 3w. After that, I transitioned from using baselines in vCenter to image-based lifecycle management in order to upgrade to ESXi 8.x.

Based on the documentation (which seemed a bit conflicting), I understood that Connection Manager should not be installed during the ESXi upgrade. So before upgrading, I removed NCM from all hosts, then proceeded with the ESXi 8.x upgrade using an image that included the Dell add-on.

Once all hosts were successfully upgraded, I added Connection Manager back into the image as a component and attempted remediation. The remediation task reported a failure, but after SSH’ing into the host, I confirmed that NCM was actually installed.

I then manually rebooted the host a couple of times (thats what you're supposed to do after installing NCM) and ran a compliance check against the image. At that point, everything showed as compliant.