Practical examples of Super Metric use cases?

[MattG]

I am looking for some practical Super Metric examples.

Thanks,

-MattG

[ServiceOptimi]

MattG- wish vmware posted a best practice guide on super metrics but they really dont have one. I have attached a doc by a vmware se for your review. WE NEED MORE SM docs... and please....not just on vsphere.

[vkaranam]

Hello Matt-G,

Please see the below link for Super metrics usage. Check out bothe part 1 and part 2 .

http://www.batchworks.de/using-super-metrics-to-monitor-cpu-ready-part1/

Thanks

VK

[admin]

Hi,

I have a few on my Blog also: www.virtualclouds.co.za

Hugo

[MattG]

Applications and the associated Tiers only have vCops System Generated metrics.   If I wanted to see Total Disk Latency for VMs in a App or App Tier,  would this work?  If so,  would this be a practical use of Super Metrics?

Thanks,

-MattG

[admin]

Hi Matt,

I havd a quick look and it should work.

Regrading more practical use cases...I am building up a few use cases and will blog around it. I have noted yorus down as a example.

Other use cases would be to use the vCD adaptor and using the vCD construct to create super metric.

[MattG]

Hugo,

Just tried it with an App Tier and it works. I will create for the Application using the looping next.

Thanks,

-MattG

[DLA_Mike]

The Best use of a super metric is across a tier or an App.  The app may be too broad depending on how many VM's but an app tier is a good one. 

CPU AVG, etc would be good across these tiers.  Also stats for the cluster where you cant get the metric natively would be another good use.

Mike

[Iwan_Rahabok]

Hi Matt,

Here is an example. This resonates well with my customers:

You have convinced your CIO to virtualise the remaining 50% of the servers.

Your CIO needs you to prove, supported by performance charts, that the platform has served every VM well, meeting the SLA in the past 1 quarter.

• Tier 1 cluster SLA: 2% CPU Contention, 0 RAM Contention, 10 ms disk latency, 0 drop packets.
• Tier 2 cluster SLA: 4% CPU Contention, 5% RAM Contention, 20 ms disk latency, 0 drop packets.
• Tier 3 cluster SLA: 6% CPU Contention, 10% RAM Contention, 30 ms disk latency, 0 drop packets.

You have 500 VM on 50 ESXi, 8 clusters, 40 datastores, 5 RDM.

You must prove that:

• Not a single Tier 1 VM has >2% CPU Contention in the past 1 quarter. The underlying ESXi also has <2% CPU contention.
• Not a single Tier 1 VM has >10 ms disk latency in the past 1 quarter. The underlying ESXi also has <10 ms disk latency.
• Etc, for each Tier and each component (CPU, RAM, Disk, Net)

How would you deliver that?

To deliver the above, you need to create quite a number of super metric. One would be Max (Max (VM CPU Contention), Max (Host CPU Contention).

You can either apply them at the cluster level or datacenter level.

Hope that helps!

e1

[vSohill]

Hi Iwan RahabokIwan

If I want to create a super metric Tier 1 cluster SLA: 2% CPU Contention, 0 RAM Contention, 10 ms disk latency, 0 drop packets. as you wrote. Can you please write thissuper metric.

If I wrote max(${adapterkind=VMWARE, resourcekind=VirtualMachine, attribute=cpu|usage_average, depth=2}) where can I put the 2% ?

Thank you

[sxnxr]

All the SLA super metric has in it is the SLA you want example 2

With that in the SM you can add it to a trend widget to have a consent line at 2. That is the only way to get a custom flat line in a trend graph

The below is using a SM just with 1 in it to provide the purple line

This is what i mean

Check out Iwan's post Operationalize SDDC program