Jump to content
Welcome to our new Citrix community!
  • 0

Other than C Drive writes, what can cause PVS write cache to fill up?


Phil Johnson1709161405

Question

In the past, I have been able to use procmon to determine what is causing our VDIs' write cache to fill up.  This time, monitoring C drive writes is not adequately identifying the culprit.  For example, I will capture about 600MB of writes to the C drive but will see 5GB of cache consumed.  Are there other tools I can use to identify what is consuming our write cache, and if so, what do I need to look for?  We use Device RAM overflow on local hard drive, on a fully patched Windows 10 image running CVAD 2203 CU3.  Note.  I have ruled out our A/V solution as the cause in this particular case.

Link to comment

4 answers to this question

Recommended Posts

  • 0

Windows Updates and software micro updates usually fill a lot of that.

 

I have taken to making the Writecache equal to the size of the c:\ drive, but thin provisioned.  I also have been increasing the size of the Ram cache for multi-session machines as the price of ram has been dropping.  This drastically improves application launches and reduces iops.

Link to comment
  • 0

Check scheduled tasks.  Even with Windows Updates disabled I have seen scheduled tasks kick off updates outside of that.

 

Edge, Office, Teams, etc all will update themselves, even if you have the engines disabled.

 

Also, if you are using fslogix, it will cache data locally as well that will utilize a lot of caching.

 

You probably just have to increase the writecache drive.  Not many other options.

Link to comment
  • 0

This is an older post but thought I would add my 2 cents based on our experience. 

 

As mentioned above by others we audit every time we patch/update to see if any apps have enabled auto-updates. But we find that the majority of the time when our vDisks start to fill are due to Windows updates that have somehow turned on and/or needing to defrag our vDisks. 

 

windows updates

 

As there are so many ways that Microsoft tries to "self heal" windows updates it was like playing "whack-a-mole" with  scheduled tasks so we added to our script to seal our images to rid the service itself. Then we restore when we crack the images open. Here is what we remove when we seal the image. 

 

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\wuauserv]
"DependOnService"=-
"Description"=-
"DisplayName"=-
"ErrorControl"=-

 

 

Defrag the vDisk

 

If we have determined windows updates is disabled we have found that after we patch even with only one maintenance chain that the vdisk will grow to which we will then merge and defrag. 

 

We also use FSLogix and do not find much is written locally but this is our experience to which we all know mileage will vary for everyone. 

 

 

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...