Can't patch beyond VCSA 7.0.1.00300: Invalid Type,expected String, instead got NoneType

[actyler1001]

Hi everyone, I've tried to patch my lab instance of the VCSA multiple times since originally upgrading to 7.0.1.00300 and it's been rough. I must say that the VMware 6.5 and 6.7 builds were much easier to maintain thus far.

Anyway, the latest issue occurred while using the VAMI to stage the 7.0.2.00200 update first, then ran install after stage was completed.  I got this error and after about an hour of waiting, no change.

Revert snapshot, disable HA for my single cluster, disable DRS, apply "config.vcls.clusters" hack to get vCLS VMs redeployed.  Enable HA, Enable DRS, all normal again.  That was fun... NOT.

It is worth noting that my original upgrade to this version of the VSCA 7 was really rough too, I have a post about there here:
https://tylercrew.com/?p=428

Regards,
Adam Tyler

[ExanJamie]

Seeing the same error message myself in production unfortunately. Though mine's at the 93% mark from 7.0.1.00200 to 7.0.2.00500.

Creating a Sev 1 ticket with support and no response at all for 2 hours. So much for the 30 min SLA 😞

Very disappointing.

[actyler1001]

Yep, I'm willing to bet real $$ that the first couple of support resources you speak with are going to be zero help too.  Took me 4 weeks recently to get a VMware engineer on the phone capable of answering some basic questions about logging and partitions in ESXi.  Company is going to crap.....

[burgisserkevin]

Hi actyler1001,

Do you have any updates on the error "Invalid Type,expected String, instead got NoneType" ?

I am unable to update to 7.0.3.00100 and cannot find any relevant information online 😞

[arvin_j]

I just hit the same issue today going from 7.0U2 (7.0.2.00000) to 7.0U3c(7.0.3.00300).

Enterprise eh?

[actyler555]

the vSphere 7 product is garbage.  Will be interesting to see what happens when 6.5 and 6.7 go EOL later this year and tons of customers are forced to face an inferior "upgrade"..  Maybe it will force their dev teams to start actually resolving these problems.  Don't hold your breath...

[arvin_j]

Spoiler

I was finally able to push thru the update in my case.  Again I am not sure of the root cause of initial issues but this is what I did:

1. Confirm all disks had space.  My Var/log was full so i expanded the disk.

2. using the MOB utility I removed all plugins outside of VMware.

3. I kept my WCP service off as I had Tanzu projects prior on this environment.

Best,



Arvin

[Kazza81]

Same. But I was able to fix it.

During the upgrade, envoy proxy was not able to start. Reverted back to my snapshot and before I started the upgrade again, I had to change the envoy config (/etc/vmware-envoy/config.cfg).

SERVICE_DEFAULT_ARGS=(
-c /etc/vmware-envoy/config.yaml
-l $SERVICE_LOG_LEVEL
--log-format "$SERVICE_LOG_FORMAT"
--service-cluster vmw_cluster
--drain-time-s 60
--disable-hot-restart
--bootstrap-version 3
)

The last line "bootstrap-version 3" is essential.

Maybe this helps others for the upgrade.

br,

Reto

[mhinterholz]

I solved the problem with LSdoctor.