> 3) VM is still working, but soon I will run out of disk space.
Do not allow the VM to fill up the datastore - that will corrupt the data inside the latest snapshot.
If that can occur very soon - you must power off the VM as soon as possible for a couple of minutes.
In that short downtime you would either copy latest delta to another datastore or add another snapshot manually.
When that is done you can power on the VM again and relax a bit.
While the VM is running on a snapshot in another datastore the next step is to use vmkfstools -i to clone parts of the chain into new consolidated basedisk. It may not be possible to clone snap 254 to a new disk. If that does not work you would try to clone snap100 for example.
If necessary it is possible to get the situation resolved with a small number of 5 minutes downtime.
What ever you do - do not try a consolidation of the whole chain while the VM is running -the risk is way too high.
When you dont act fast enough and the datastore is already filled up - it is very important to not answer the popup question: "VM is running out of space - do you want to continue - or abort ?"
Instead leave that popup unanswered and free up space and only answer the question when additional space is available.
Ulli