Failed to restore saved state once again

[lefevrol]

I have an Ubuntu VM that I saved a while back, unfortunately without taking
a snapshot of it in Powered Off state first, and now Virtual Box, after an
intermission of a few weeks, refuses to restore it. This is with VBox 3.0.0.
I have not upgraded VBox or made any changes to it in the meantime, in fact
I have not used for a while, nor has any change been made to the *.sav file,
which is still dated 14.07, so it is totally absurd to be told in the error message
that it must have been damaged or from an older version of VBox. Now I lost
all the work I did on and with that VM. Neither is this the first time that
something like this happens. Frankly there is no excuse for such flakiness
in a version 3.0 software. VBox is nice when it condescends to work but,
in my experience, it's just unreliable: a toy that you cannot risk depending
on for real work. Will it ever shape up, I wonder.

[lefevrol]

The problem seems related to shared folders as I find in the log:

00:00:03.304 SharedFolders host service: connected, u32ClientID = 10
00:00:03.304 SharedFolders host service: loading state, u32ClientID = 10
00:00:03.304 SharedFolders host service: loading folder [C:\products\GO]
00:00:03.304 SharedFolders host service: unexpected saved name length 24, should be 34
00:00:03.304 SSM: LoadExec failed with rc=VERR_SSM_UNEXPECTED_DATA for unit 'HGCM'!
00:00:03.304 Changing the VM state from 'LOADING' to 'LOAD_FAILURE'.

A problem with shared folders, which are totally secondary, strikes me as
an extremely flimsy excuse for aborting a launch and thus a prime case of
unsound judgment and poor coding (as in: "I'm going to bail out at the first
provocation; heaven forbid I should try to make the launch procedure robust!")
but let's ignore this for now. I removed all shared folder entries from the XML
config and now I get:

00:00:03.393 SharedFolders host service: connected, u32ClientID = 10
00:00:03.393 SharedFolders host service: loading state, u32ClientID = 10
00:00:03.393 SharedFolders host service: unexpected saved state 1, should be 0
00:00:03.393 SSM: LoadExec failed with rc=VERR_SSM_UNEXPECTED_DATA for unit 'HGCM'!
00:00:03.393 Changing the VM state from 'LOADING' to 'LOAD_FAILURE'.

What the heck is "saved state 1/0" and what, if anything, can I try now?
Neither Google nor the search engine on the VBox web site were of any help.

[Ralph1]

Hello,
i am no special expert, however saving a state means saving everything together
and this seems to be broken by you removing the shared folders
the 1/0 could be a flag-bit telling some component to load the shared folder
state which is now missing ...

however if you recently did a backup ... the chance to use a backup of *all* the
files of that vm taken before the failure occured is much better than trying to
reconfigure a saved vm ...( because it will become inconsistent if some saved
feature is missing during restore like you know already ... )

even if the restore could/would ignore the inconsistency made by the changes
who cold tell what would happen by restoring data/memory and so on
wich belongs to a 'switched-off' feature ...

Ralph

[lefevrol]

Ralph, all the shared folders referenced in the XML config still existed, so there was no reason for these errors. I removed them from the config only because it looked like the most sensible thing to do in the face of those errors. Anyway, I moved to VMWare: its price is now much less than I remembered and my time is worth that.