Number of vCPUs left in the Cluster

[Souad90]

Hello Everyone,

I have been working in a dashboard of capacity management, and I want to show how many vCPUs are free in the clusters. Is there any way to get that ? I looked in the metrics list and I can't find any one that gives the value I'm looking for.

Thank you.

[sxnxr]

You don't need a super metric as it should already be there.

Based on you wanting the remaining vcpus i am guessing you are using a allocation based capacity model and not a demand. There are 3 useful metric

CPU -> Allocation -> Total Capacity (vCPUs). This will give you the total vCPUs that are available to your cluster. It will also take into account any overcommit ratieos you have in your policies

CPU -> Allocation -> Usable Capacity. This will give you a total of the usable vCPUs in your cluster. It removes and HA\Buffers % that you have in you policies

CPU -> Allocation -> Capacity Remaining (vCPUs) This is the one you are after. This will take all the VM's allocated vCPU's away from the usable capacity and show you how many vCPUs you have available for allocation in your cluster.

[Souad90]

Thank you for your replying. I have tried this metric in a view and it is not giving any result, what I get is a dash : "-". I don't know why!

In fact I want to display it in a Scoreboard so I made an XML file and the result is not dash but an interrogation mark : "?". Can you please try on your side and tell me if you get a value or not? Thank you.

[sxnxr]

Check to make sure the metric is enabled in you policy If you have more than 1 policy make sure it is enabled in all of them

[Souad90]

The metric is already enabled it's a simple one not a super metric.

[Souad90]

Why this metric : Capacity Remaining value (vCPUs) is not working ?

[sxnxr]

Not sure. The only 3 times this had not showed anything is

1      Using a demand based capacity model

          If you are using demand in your policy this metric will not work.

2      Disabled in the policy

          Regardless if it is not a supper metric it can still be disabled in the policy. Double check to make sure it is enabled for all object types (See attached)

3      Rebooted the entire cluster (this includes software updates)

          It can take up to 24 hours for this metric to be worked out after a reboot.

[Souad90]

Thank you a lot. I'll check and let you know. Could you please explain to me what are the two columns KPI and DT mean in the table ? I'm always wondering what is their purpose. Sould they also be enabled?

[sxnxr]

It depends if you want to use the metric for them.

The KPI determines whether the metric, property, or super metric attribute is considered to be a key performance indicator (KPI) when vrops reports the collected data in the dashboards.

DT is better explained here vRealize Operations Manager 6.0.1 Documentation Center

[SWARob]

Running vROPs 6.2

I'm also having a similar issue.  Trying to get the total vCPUs for a cluster.  Results are coming up blank.  Why would my KPI and DT be "not" inherited?  I tried setting them to "yes" and still no results.

Many thanks in advance!

Robby

[sxnxr]

Check that you have enabled the allocation based model in the policy on the object that you want to model on and also the ratios for over allocation. This is what it uses to calculate the vCPU allocation stats. Total, Usable remaining etc

The KPI and the DT are not used much for this metric. DT for example, this metric is not dynamic for alerting in the way CPU usage is so having a dynamic threshold is not worth wile