How to delete SnapShots with least amount of extra disk space

Hello,
does anyone know how to delete SnapShots with the least amount of extra disk space needed (if the disk got nearly full do to SnapShot's)?
I understand, that a SnapShot takes up disk space and in order to delete a SnapShot the data of the SnapShot needs fist to be copied into the virtual disk or another SnapShot and after that the files representing the SnapShot can be deleted to free up disk space. So in general you first need disk space to free disk space. My problem is that the SnapShots for TimeMachine run well for many weeks, but suddenly the bloat up and fill the disk quickly.

My question now is: how can this be done in a way that uses the least amount of extra disk space?

Consider you have 4 SnapShots and the current state:
Snap0 -> snap1 -> snap2 -> snap 3 -> current state

In what order do I delete the Snap's to use the least amount of extra disk space? Or asked differently: in what direction will the data of a SnapShot be moved and when will the datafilee from the Snap be deleted? If I delete Snap3 does the data go into the current state or snap2 - and will the datafile for this snap be deleted after completion?

Should I delete snap3, snap2, snap1 and last snap0? Or the other way around?

I have tried different things but did not understand why sometimes the extra disk space needed is minimal and at other times not. Today I deleted 3 Snap's and the disk was filling up more and more. During deletion I got the msg that no new snap could be generated and around that time the used disk space started to decrease dramatically .. feeing up 100GB of space. So maybe the snap datafilee only get deleted when a new SnapShot is tried to be generated?
Any insight is wellcome.

 

Thank you with your step by step guide. I have tried it and it didn't work as I was thinking it would.
I had 3 Snaps, the representing files of the Snaps took around 4Gb in size. I started deleting the oldest one and worked my way to the newest one. The available disk space went from 54GB to 45GB to 35GB to 24GB - it took around 30GB in additional space to delete Snaps of size 4GB. Why is that?
After all Snaps where gone I could free up unused disk space of the Virtual machine of 15GB but that didn't do much in freeing up the actual physical disk space.
Only after 2 restarts of the host machine the disk space was released and suddenly I have 100GB free space. A jump of 50GB I have no idea where they came from.
This drives me nuts. It is completely unpredictable how much disk space the virtual machine uses and how to free space when the physical disk gets full.
You got maybe some more insights about how the virtual machine clams disk space of the host system and how it frees it?
Thanks a lot.

 

So far I have not found a way to reliable predict what happens when I delete Snaps. I tried newest first, oldest first, middle first. Sometimes it did not take much additional disk space deleting them, some times it took so much, that the disk of the host system ran out of space.
What I don't understand is: Why does Parallels claim so much additional disk space when Snaps are deleted, even if the content of the virtual disk only has changes minimal?
It would be really cool if someone with inside knowledge could share what is happening with disk space if you use/delete Snaps.

 

And after a few days of automatic Snaps (TimeMachine optimized Snaps), the available disk space of the host system jumped from 100GB to 160GB. All this despite the fact, that the VM has hardly changed. This is really some very strange allocation and dislocation of disc space in the host system. I wished someone could shed some light on this. I hate it when I have no idea what is going on and can't predict if the VM can be left alone for extended periods of time or if one has the check daily what the hell is going on, to make sure the VM does not run out of host disk space. It has happened to me before and resulted in a VM that could not be recovered. Wich was really really bad.

 

Thank you Susan for taking the time to respond. This is exactly what I assumed I should do, because it felt it would make technical sense. Problem is: this is NOT working as expected. More often than not this deletion process needs excessive additional disk space and one time while experimenting I have damaged a VM volume beyond repair. If this would have been a production system this would have been a complete disaster.
Since SmartGuart is doing optimized SnapShots for TimeMachine I am always worried that this might fill suddenly the host system volume and I end up with a damages VM Volume that I am unable to repair.

 

Hello Susan,
way do you repost your old post, when I just told you that your advice is NOT working? Does anyone else have some real insight how this works?
Thank you.

 

thank you John for responding. But the host system is a Mac running MacOS Catalina (MacOS 10.15, APFS disk format) not AWS or other cloud services.
Non of the above mentioned methods work and deleting Snaps in Parallels for MacOS is completely unpredictable regarding disk usage ...
I hoped someone might know something regarding disk usage under MacOS host system. The virtual machine runs UBUNTU 20.04 ... not sure if that makes any difference.

 

Hello John,
why do you repost your old post, when I just told you that your advice is NOT working? Does anyone else have some real insight how this works?
Thank you.

PS: Are these post from bot's that need to have the last word and that just post some nonsense?

 

Hello Harry,
Are you sure about this? Why would the system sometimes put a Snap into the current state and sometimes into an older Snap? At random? Does not make any sense to me. And yes if you have enough disk space this is not an issue. But I would like to know what to do when it is in fact an issue and you have limited disk space available. How come that no one seems to know exactly what is going on. All answers so far assume and guess what's going on. No one know for a fact and can predict what's going to happen when you delete a Snap.