> don't know about the issue with encryption, but in this thread and elsewhere people are talking about "shrink" and command line options.
Shrink and compact seem to both be used interchangeably at times. What I am trying to accomplish is to remove 'unused' space on the host for a Virtual machines .vmdk files.
I don't use 'pre-allocated' so the used Host space increases as you add more data and often even if you add and then remove data for the same net currently allocated size. (Stuff presumably gets marked as used and Vmware apparently isn't aware that it is no longer in-use when it is released).
> I like to remind that it has "always" had Compact in GUI.
The Workstation 17.6.3 GUI only seems to have the 'clean up disks' option for a Windows Guest. I do not see this for Linux Guests.
My concern about Encrypted Guests was that (at least at some point in the past) the Linux instructions had you do a manual step of writing 'zeros' to all available space in the guest. (Followed by a Host step to actually compact the disk with a command line now that it had been prepared with zeroes).
In the past, I've had encrypted Linux Guests react to the 'write zeroes' step in the guest by tremendously growing the Hosts used drive space for the guest. (Which did not seem to happen for un-encrypted Linux Guests). I assumed this was because the 'zeroes' got written to the .vmdk's as an encrypted version.
I hadn't tried this in a very long time, so I wasn't sure if this was still an issue or not.
As indicated in my subsequent post, I was able to try the (new to me) method of performing the command
sudo vmware-toolbox-cmd disk shrink /home
from the Linux Guest and it seemed to eventually work without 'blimping out' the Host space beyond what you would expect for an un-encrypted guest.
It just needed the amount of space for an individual part of the multi-part (in my case) .vmdk files as it worked through them.
I did get a warning about
kernel:watchdog: BUG: soft lockup - CPU#0 stuck for 127s! [kwin_wayland:2221]
I assume this was because some type of Guest processing must have gotten stalled during the process.
I'm not sure if that can safely be ignored or not.