vcenter unable to create vsan alarm because it allready exists

[Darking]

Hi.

after upgrading to 6.7 U3 we are having an reoccuring alarm creation error.

the error message says:

The name 'alarm.vsan.health.test.cloudhealth.v1diskcheck' already exists.

i havnt been able to figure out what it is referencing, but im dont see any immediate errors anywhere in the vsan health view either.

Any ideas how to get rid of it?

[TheBobkin]

Hello Darking  

The Health alert it is referencing is likely on-disk format version checker.

Looks like you are not alone in encountering this:

vCenter try to create a task then give "the name already exists"?

But that being said, I have not seen this issue in GSS (looks to be more  vC-related and also not something that would come in as a P1 so maybe just hasn't passed my desk), I'll have an internal look when I get the opportunity.

Everything else went okay when upgrading the vC and nothing else out of the ordinary?

Bob

[Darking]

Hi TheBobkin​

The upgrade experience going from 6.7 U1a to U3 went very smoothly and we have not seen any issues in a month since the upgrade, except that error that pops up often.

It took a few hours for our 8 node stretched cluster, but nothing to worry about

ive checked the ondisk format health status and it reports in 0 disks with older format, so there has definately not been an error there.

we are running build 15160138 on the hosts and vcenter is 15129973

the biggest issue is that i dont even know what the alarm is referencing really. but if you say some sort of on-disk alarm, that makes little sense.

[TheBobkin]

Hello Darking​,

"the biggest issue is that i dont even know what the alarm is referencing really. but if you say some sort of on-disk alarm, that makes little sense."

So, the thing to understand is that it is not alerting over this health test failing/warning but the actual vCenter side configuration of this alarm itself encountered a problem - I have looked into this internally and it does appear to be a known issue, specifically with the de-registration of a deprecated alarm entry that got replaced.

In the reported occurrences of this particular alarm (v1diskcheck) there is indication that this stops re-occurring following restarting vsan-health service on the vCenter - code changes to prevent this occurring in future have been submitted but I can't provide a data/version.

Bob

[EricRS96]

I have the same problem since October 2019. I opened a ticket to VMware, they check and didn't have an answer.

I found the alarm was set in the top of the tree (vCenter), so I deleted the alarm. After a while it recreated itslef and the message was gone. So I close the ticket.

But since the december update, it came back. I deleted again the alarm, then it created itself but the message is still there after, so I just ignored it.