Custom Dashboard trying to make a clean health tree

[AEsansun2]

I am creating a custom health tree so my Help desk can see any issues.

The problem is , right now i have 4 vm's that are yellow. I double click on the vm that's yellow and i still don't see the issue.

If i click on the red icon , it gives me a list of itmes Badge|health , Badge"Risk..

is there anyway in plain english to tell me the issue . I don't really care about badge - all i really care about is current issue..

n

[mark_j]

In the Custom U, the default value "Health" is calculated using anomalies and faults (using KPIs to highlight key metrics), so you're going to want to attached an alert widget or root cause ranking widget to see a list of anomalies and faults. "Health" color (state) is a value that is adjustable in the Custom UI global preferences if you wish to mask this visual representation of the indicator.

However, in the vSphere UI a VMware vCenter Adapter derived metric called badge|health is created and used for the default health value.. this considers things like workload in addition to anomalies and faults.

The concept of derived metrics is leveraged for other adapters in the Custom UI, though, such as those included with Hyperic. The Hyperic MP allow configurable derived metric such as workload.

#vC Ops 5.x

[AEsansun2]

Thank you Markj.

So be more specific; I am using the custom dashboard and have a root cause ranking widget. When you click on it all of them are DT issues.

As far as the help desk is concerned that is not something they really care about.  They only want real critical issues .. otherwise they want it all green.

I tested by following this guide to turn off DT   - http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=204000...

thinking that now it will not report on DT ,  But then everything instead of being yellow turned green. So I guess that doesn't work

I then tried adjusting the Health Status in "Global Status" and it seems to have worked.

I put bad from 0 to 1 .. etc. This way I am expecting if there are any real issues (not DT) that i will be alerted. Is that correct ?

[mark_j]

Well, no. DTs should always be running and you should never turn them off. Turning DTs off would cripple vC Ops's ability to determine anomalies and subsequently diminish it's value as a solution.

Changing the global health colors controls that coloring, but isn't going to control the calculation of the scores. When there are real issues, they will impact the health score just like anything else (anomalies are blind), however you can weight metric using KPIs to impact health scores more if needed. If you have the colors dulled and am relying on them for indication of a problem , you will fail to see an issue. However, you could have alerting set up with KPIs for items that you feel are legit.. which can be filtered to using the alerts widget and/or filter rules in emailFilter.xml.

[AEsansun2]

OK, so i have DT running now. I thought even if the colors are dulled ; if there is a real issue it would show up.

example if a vm goes down, would i see it even though the colors are dulled. Or vm going down just contributes to the

health score ?

[mark_j]

Depends what you mean by "down". If you mean powered off, it would stop collecting data for that VM and it'd show at BLUE with "?"s for badges rely on life values.. this includes the health score.