This post is more than 5 years old

145321

January 23rd, 2012 21:00

Over-sized VMs

Is there an easy way to generate a report of VMs that are over-sized (too many CPUs, too much memory, too much disk space allocated)?  I thought I remembered seeing that in some of the demos, but I can't find any references to it in the docs I've checked. If it is in the docs somewhere, a reference would be great!

Thanks

Steve

January 24th, 2012 22:00

If you are using the latest version the reports were renamed to ...

Virtual Machines Memory Addition - for VMs that have less than X % overallocated memory.

Virtual Machines Memory Reclamation - For VMs that have more than X % overallocated memory.

On the left Panel under homes, choose the reports dashboard, you can find it in the list of templates.

Golan

January 24th, 2012 17:00

For memory you have 2 reports

Virtual Machines with X% or less Overallocated memory

Virtual Machines with X% or more Overallocated memory

There is also a rule/alarm called  VMW Virtual Machine Zero Page Memory Detection.

For CPU the best may be to create a custom report/dashboard that just has a table of ALL virtual machines with columns for name, vcpus, CPU usage over a time frame, sorted by CPU ascending to decide if it is over allocated.

Golan

January 24th, 2012 20:00

Sorry -- can you tell me specifically how to access these reports?  They sound like what I'm looking for, but I don't see them anywhere.

January 25th, 2012 14:00

Thanks -- still learning the interface, and where to find what I want.  I appreciate the help!

21 Posts

January 27th, 2012 18:00

Hi Golan,

Sorry to hijack the post. But I was wondering, I ran the "reclamation" reports and it found only 2 VMs in our environment. I do know for a fact that we have couple 2008 R2 VMs with 16GB of RAM and when I look at them using VCenter Performance or vFoglight vmexplorer, the memory usage is roughly at 4% to 6%. I was expecting those VM to appear on the report but they did not.

Our vFoglight implementation was done Tuestday Jan24th 2012, maybe we need to gather more metric for a longer period?

What are you inputs on that?

Thank you

Alex

73 Posts

January 30th, 2012 14:00

sbc350,

When you ran the report did you select a service that included all VM´s in your environment (i.e. your Virtual Center) and what time frame did you select?

In order to get accurate results from a performance perspective I would wait at least 30 days to let it gather more performance data about the environment.

For testing though you can try set time range to last 4 hours just to see you get some result back.

/Mattias

No Events found!

Top