I may have a hang on my attempt to delete a snapshot, but not sure. I just upgraded to Parallels 4.0 and realized I still had a snapshot from the earlier version, so I attempted to delete it. Well the UI froze and the beachball appeared, but something has been happening. I opened a terminal window and have been watching the files instead the guest package. The .hds file grew and a smaller one eventually disappeared. The parallels.log file logged... 13:08:47.242 F /vm:31395:B0081000/ Vm {7d51ac48-0a63-413c-b5c9-52389ffe8cf9} processing command vmCmdDeleteSnapshot .... 13:08:47.383 F /LocalDevices:31395:B0081000/ [AppHDD] Disconnect 13:08:48.333 F /vm:31395:B0081000/ Vm {7d51ac48-0a63-413c-b5c9-52389ffe8cf9} processed command vmCmdDeleteSnapshot with status 1 14:24:59.978 F /vdisk:31395:B0081000/ Destructing when the disks in the list not released 14:25:00.238 F /vdisk:31395:B0081000/ Close: wait for thread completion 14:25:00.583 F /vm:31395:B0081000/ [Profile] IDE HDD 0 creation time is 52 msecs That was 2.5 hours ago. But disk file gets its timestamp updated every few minutes. rwxr--r-- 1 weave staff 15045182464 Mar 22 16:56 disk2.hdd.0.{5fbaabe3-6958-40ff-92a7-860e329aab41}.hds ... and is growing very slowly. Fortunately I made a backup of the 3.x folder with the guest in it before upgrading, so worst case I can restore that if this is corrupted, but figured I'd post and ask about this anyway, in case someone else runs across this. If the issue here is snapshots should be deleted before an upgrade, it'd be nice if the upgrade procedure checked for that and threw a warning before continuing. It'd also be nice if the UI would do *something* besides just hang and show "application not responding." I still think this may complete eventually, so I'm letting it run -- for now.
I killed it after nine hours, but somehow it worked anyway. The snapshot was gone when I restarted Parallels and I was able to do the compressor. It triggered a report to Apple, which I filled out with the above log entries. Funny thing is, when I ran the compressor, the disk compression bit crashed at the end, generating an error in Windows, which I also "sent to Microsoft" --- but the disk compression appeared to work because it freed up six gigs.
