Hello:

​

I am on Workstation Pro 17.0.0 with a collection of about 10 VMs that need to by up near constantly for medical reasons. I have searched for the most optimal way to maintain the disks used by these VMs and was told to do things like disk cleanup, clear temp files, and defrag then run a utility like sdelete (sysinternals) to prepare the VMDKs for compacting before backing them up. I've put all of that into a batch file and set each VM to fire the batch file about 1 hour 45 to 2 hours after the last one beginning on Friday night at around 23:00 and working all the way up until the last one fires at about 13:45 on Saturday.

​

well, this is now the second week that I've had to manually intervene as of 19:50 all of the VMs were still running the weekly maintenance script which was causing a global slowdown on my host laptop with 32 GB of RAM.

​

The specifics of the setup that are common to all VMs is as follows:

1) All are using dynamically allocated disks (2 disks each) split into multiple files. This was chosen because of storage constraints and financial constraints.

2) Everything is on a USB external HDD (WD My Passport 4 TB) This was chosen at the time because it was not only within budget but also had a lot of space to allow me to grow as my needs changed.

3) Host OS: Windows 11 x64

4) Host machine RAM: 32 GB

5) All VMs set to use about 512 MB each

​

the weekly batch file is below for reference

@echo off
title Windows Maintenance
rem delete files in %TEMP% and subfolders to prepare for defrag
call DEL /f /s /Q %TEMP%\*.*
rem run disk cleanup and clear all available file types on C:
call cleanmgr /d C: /sagerun:1
rem defrag each partition on each drive
call defrag C: /f
call defrag D: /f
call defrag E: /f
rem run sdelete to zero out all free space on all partitions of all drives to prepare for compacting
rem this should (hopefully) be done manually through workstation within 24-48 hours of the scheduled job completing 
call sdelete /z /c C:
call sdelete /z /c D:
call sdelete /z /c E:

​

What I've noticed is that even with the gap between when one VM starts the above maintenance script and the other starts it (about 2 hours max) somehow things seem to get contentions on the USB disk where the VMs are stored which is seemingly the bottleneck in all of this.

​

Short of moving to an external 4 TB SSD (which is outside of my budget at the moment) are there any suggestions that can be given me to 1) minimize the impact this has on my operations weekly? 2) create less wear on my HDD with about 10 VMs all trying to write 0s to empty space at the same time?, and 3) make the weekly maintenance a lot smoother of a process. If anyone has any recommendations for a better program instead of sdelete or suggestions on how to improve my maintenance script or any thoughts on any of it, please let me know as I will look into it. Right now, physically purchasing anything (e.g. SSD, software, etc.) is a bit out of the question due to budgetary constraints so solutions requiring purchasing anything will be put on the back burner and free solutions will be prioritized.

​

Any help in this would be greatly appreciated and thanks for your help.