VMware Cloud Community
TristanT
Contributor
Contributor

Frustrating alarm behavior in vC 4.1

Here's an example: at top level of vcenter, I set the "Datastore Overprovisioning" alarm to Trigger a warning over 150% and an Alert over 200%. Action is to send an email to our vSphere admins.

Any generated alarms come across as follows:

Target: USPV_2_3B80_02
+ Previous Status: Gray+
+ New Status: Red+
+ Alarm Definition:+
+ ()+
+ Current values for metric/state:+
+ Metric Space potentially used = 25,063KB+
+ Description:+
+ Alarm 'Datastore Overprovisioning' on USPV_2_3B80_02 changed from Gray to Red+

So the alarm definition should be "Yellow Metric is above 150%; Red Metric is above 200%".

The Current values should read "Metric Space potentially used = 250.63%"

So why does the alarm come across as percentage * 100 in KB? I can interpret this alarm for what it is, but it is pretty darned confusing to everyone else on my team.

Anyone else observed this in vCenter 4.x?

Anyone have an idea on how to fix this?

Thanks much! Figured I'd check with community before filing a support case.

Reply
0 Kudos
8 Replies
Neuroflux
Contributor
Contributor

Yes, I have the same thing with Datastore Usage (Datacenter level).

VC 4.1.0 Build 258902

This should be 90% and 95% for the trigger. Metric Space is 98.19%

I haven't done any other alarms yet so not sure how deep this runs.

=====================

Target: Storage-33

Previous Status: Gray

New Status: Red

Alarm Definition:

()

Current values for metric/state:

Metric Space actually used = 9,819KB

Description:

Alarm 'SSI> Datastore Usage' on Storage-33 changed from Gray to Red

Reply
0 Kudos
Ashley004
Contributor
Contributor

Hi

Did you resolve this or find out anything regarding the metric now using KB insted of a spcified percentage ?

Reply
0 Kudos
jbiggley
Contributor
Contributor

I ran into problems with memory and CPU usage on guest machines. False alarms triggered. I disabled them this morning as the frontline support team was threatening to storm my office because of the false emails. We'll just monitor the VMs other ways, but wondering if anyone else has had this problem in addition to the datastore problem?

Reply
0 Kudos
Neuroflux
Contributor
Contributor

Not yet. It's not really a huge issue, just annoying as the OP noted.

Most likely a coding mistake that will get fixed in the next release.

Reply
0 Kudos
khoe
Contributor
Contributor

Hi

Did anyone solve the problem with alarms shown as KB instead of percent?

-


Best Regards

Kim Hørsted

-------------------------------------- Best Regards Kim Hørsted
Reply
0 Kudos
Ashley004
Contributor
Contributor

Hi Khoe

Havent got it sorted yet , have logged a call with Vmware will feedback when i have more info

Reply
0 Kudos
mheffler
Contributor
Contributor

Ashley,

     Did you ever get a reponse to your logged call on when it might be fixed? Could I have your ticket number on this?

Thanks,

Max

Reply
0 Kudos
Ashley004
Contributor
Contributor

Hi Max

We eventually did get some feedback , that apparently the issue would be resolved in update 1 which we have not implemented just yet so I can’t verify if the issue has been resolved yet

Ashley

Reply
0 Kudos