Memory leak - local disk drives (share redirection)

P

PatrickD8

Member
Hi,
I have found a memory leak, which occurs regardless of the system used. This arises when going through directory structures of redirected network shares.
This usually ensures a high CPU utilization. Per visited directory, the memory usage of explorer.exe increases by almost 1MB.
In the worst case, this causes a user's session to freeze.

Tested:
Windows Server 2012 R2, 2016, 2019
User Profile Disks enabled / disabled

I'm not sure if it's a Windows or some Parallels agent problem. Maybe someone has an idea how to solve the problem.

Best regards
Patrick
 
Hello @PatrickD8
My name is Valentin, I lead Parallels Commercial Support Team. We would like to take a closer look at the behavior you are describing. Would you please consider cooperation with our support team so we can try to reproduce the issue locally, or troubleshoot it on your environment if possible?
Would appreciate if you could send me your contact email for further discussion in scope of support request.
Thank you in advance!
 
Hi Valentin,
Thanks for the fast respond. A short tutorial on how to reproduce the problem under Windows:

1) Mount a network share on your local machine, e.g. Test (\\server01\somedir$) (S:)
2) Enable local device and resource redirection in Parallels Client (Connection Properties >> Local resources >> Disk drives)
3) Connect to a published Windows Server (Tested on RD Session Hosts)
4) Open explorer.exe and some tools for monitoring (e.g. taskmgr.exe)
5) Open redirected drive S:
6) Browse through folder structure

If you quickly change folders, the CPU load increases for a short time. The memory usage increases constantly, the more often you change the directories on the redirected drive. The garbage collector does not clean up after closing UI.

I just tested with mstsc.exe and had the same problem there. Seems to be a Windows problem.
Are you still interested in seeing the problem in my environment?

Best regards
Patrick
 
@PatrickD8
Many thanks for detailed replication scenario. For now we will proceed with internal reproduction to observe and analyze the behavior. I'll keep you posted on anything noteworthy that we will find.
Greatly appreciate your cooperation!
 
You must log in or register to reply here.
Back
Top