-
- Novice
- Posts: 7
- Liked: never
- Joined: Jan 20, 2020 10:04 am
- Full Name: Paul Webb
- Contact:
Confusion over Memory "Demand" counter
We run a Hyper-V cluster and I am noticing that the memory Demand counter does not seem to produce realistic readings.
For example....
We have a Windows 2019 RDS server that is configured to use dynamic ram between 2GB and 32GB.
In Hyper-V Cluster Manager, I can see Assigned RAM shows 28GB and Memory Demand is 24GB
Veeam One reports the "Memory Usage" correctly, however, the "Memory Demand" counter only shows around 1.5GB. This counter also seems to be used when sizing the "Oversized VM" reports which means wildly inaccurate recommendations.
Can anyone shed any light on what this counter is and why it is so far out?
For example....
We have a Windows 2019 RDS server that is configured to use dynamic ram between 2GB and 32GB.
In Hyper-V Cluster Manager, I can see Assigned RAM shows 28GB and Memory Demand is 24GB
Veeam One reports the "Memory Usage" correctly, however, the "Memory Demand" counter only shows around 1.5GB. This counter also seems to be used when sizing the "Oversized VM" reports which means wildly inaccurate recommendations.
Can anyone shed any light on what this counter is and why it is so far out?
-
- Veteran
- Posts: 7328
- Liked: 781 times
- Joined: May 21, 2014 11:03 am
- Full Name: Nikita Shestakov
- Location: Prague
- Contact:
Re: Confusion over Memory "Demand" counter
Hello Paul,
Veeam ONE takes performance counters from Hyper-V hosts.
Could you specify what metrics are you operating with in ONE?
There is "Memory Demand" for Hyper-V VMs that is Amount of memory a VM requires to run all active processes.
And "Committed Memory" on the host-level that shows how much memory were allocated for processes and to which processes the OS has committed a RAM page frame or a page slot in the pagefile (or both).
Thanks
Veeam ONE takes performance counters from Hyper-V hosts.
Could you specify what metrics are you operating with in ONE?
There is "Memory Demand" for Hyper-V VMs that is Amount of memory a VM requires to run all active processes.
And "Committed Memory" on the host-level that shows how much memory were allocated for processes and to which processes the OS has committed a RAM page frame or a page slot in the pagefile (or both).
Thanks
-
- Novice
- Posts: 7
- Liked: never
- Joined: Jan 20, 2020 10:04 am
- Full Name: Paul Webb
- Contact:
Re: Confusion over Memory "Demand" counter
If "Memory Demand" is indeed a measure of the amount of memory required to run all active processes, then this measure is completely out.
The link below shows a typical RDS server in our environment with 32GB of dynamic ram assigned. 30GB of this is full in use within the guest OS, but the "Memory Demand" measure shows the demand to be 1.62GB. This is happening on multiple VMs across all 6 host servers.
https://drive.google.com/open?id=1LTlc4 ... Y_HrW2-u-b
The link below shows a typical RDS server in our environment with 32GB of dynamic ram assigned. 30GB of this is full in use within the guest OS, but the "Memory Demand" measure shows the demand to be 1.62GB. This is happening on multiple VMs across all 6 host servers.
https://drive.google.com/open?id=1LTlc4 ... Y_HrW2-u-b
-
- Veteran
- Posts: 7328
- Liked: 781 times
- Joined: May 21, 2014 11:03 am
- Full Name: Nikita Shestakov
- Location: Prague
- Contact:
Re: Confusion over Memory "Demand" counter
Yes, that is the Microsoft's definition.
Could you re-share the screenshot as it's cannot be opened?
Could you re-share the screenshot as it's cannot be opened?
-
- Novice
- Posts: 7
- Liked: never
- Joined: Jan 20, 2020 10:04 am
- Full Name: Paul Webb
- Contact:
-
- Veteran
- Posts: 7328
- Liked: 781 times
- Joined: May 21, 2014 11:03 am
- Full Name: Nikita Shestakov
- Location: Prague
- Contact:
Re: Confusion over Memory "Demand" counter
This link works.
However the numbers look strange.
I would suggest to check the values through Perfmon on the HV host to find out if Veeam ONE collects the metrics incorrectly or the issue is on HV side.
Or you can just contact Veeam support and make them investigate the issue.
Thanks
However the numbers look strange.
I would suggest to check the values through Perfmon on the HV host to find out if Veeam ONE collects the metrics incorrectly or the issue is on HV side.
Or you can just contact Veeam support and make them investigate the issue.
Thanks
Who is online
Users browsing this forum: No registered users and 8 guests