First I want to thank everybody who offered me help; your help is not needed anymore (unfortunately, not because the problem was resolved, but because there is nothing to recover).
Summary of what happened: ========================= <off-topic for this list, please read only if it interests you> When you shut down VMware, or it crashes, it's like shutting down a computer: you lose all the volatile information (RAM, processes, etc.). You don't lose the content of the hard disk (unless the unexpected shutdown damaged the filesystem integrity, unusual in todays' file- systems). VMware has two special features, that are not "supported" by a normal PC (I ignore OS-level features such as "stand-by"): 1. You can always suspend a running guest, and later - resume it, exactly as-it-was, the same processes, memory, windows, open files, sockets (though the other side probably closed the connections), time (!), etc. Of course, the disk looks exactly as it was when suspending, too. 2. You can always take a "snapshot" of the current state; It's exactly like "suspend", but you can have multiple snapshots. VMware organizes them in a tree (i.e. snapshot4 may be the follower of snapshot2, while snapshot5 is the follower of snapshot3, and both of snapshot2+snapshot3 are the followers of snapshot1; This also allows VMware to save only the differences between a snapshot and its "parent", rather than the entire stae for each snapshot). I use suspend-resume a lot, and from time to time - also take a snapshot. In this morning, VMware crashed. This is not so awful, because (as I wrote above) the content of the disk is kept. I restarted VMware, and everything was there, in the HD. So far so good. However, as I wrote in a previous message, this distro was a live-CD, so the main filesystem was volatile (ram-disk), and the HD was only a mounted filesystem. And I had some silly modifications to a file residing on the ram-disk which I didn't have any chance to save before VMware crashed. So instead of living with the updated HD but the not-so-updated RAM file, I insisted on recovering even this file (mistake #1), so I decided to "visit" a snapshot (that contained this updated file), copy the file to a directory outside of the VMware zone, and then running again the "normal" VMware (not a snapshot), and copy the file in. What I didn't know was that when you "go to" a snapshot, it's not like a VISIT (so later you can go back to the "trunk" of the "tree" and still have the latest HD), but VMware removes all the changes made to the hard disk since the last snapshot (my mistake #2). When I "went back" to the snapshot, VMware probably popped up a warning dialog box which told me that data may be lost by doing it. This warning dialog box was so similar to the "VMware-tools" warning dialog which pops up any time a guest machine is powered on, so I clicked on it automatically, assuming it is the "VMware-tools" dialog box (mistake #3). So now I have (hurray! ;-) the updated RAM-file (from the last snapshot), but the HD reverted to what it was in the last snapshot :-( By recovering (successfully) a stupid file, I lost all the work of the last days (20 hours at day... ;-) In the first minutes after it happened I didn't understand exactly what happened, so I hurried up to post an e-mail to the mailing list. Now I know that there was nothing that you could help. Conclusions: ============ 1. When going to a snapshot, ALWAYS take a new snapshot before (actually, I think that VMware had to do it automatically, by default, and just let you skip it by a dialog box before starting to take the new snapshot). 2. Disable the "VMware-tools" dialog, so only critical messages will be displayed next time. 3. Find a solution for VERY recent backups of VMware directory, that meets the challenge that VMware files are so huge. 4. Send an RFE to VMware, asking them to color dialog boxes which warn about lost data in a special color; It's confusing that the "VMware-tools" dialog is so similar to the "lost-data" dialog. Of course, all of that doesn't help me this time, so let's go to the "solution": Solution: ========= Well, not really a solution: just re-code everything I did in the last days, as long as it is fresh in my memory. </off-topic> Thanks again, -- Eli Marmor [EMAIL PROTECTED] Netmask (El-Mar) Internet Technologies Ltd. __________________________________________________________ Tel.: +972-9-766-1020 8 Yad-Harutzim St. Fax.: +972-9-766-1314 P.O.B. 7004 Mobile: +972-50-5237338 Kfar-Saba 44641, Israel ================================================================= To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
