[solved]Canceling a partition resize operation while in progress?
Hi, Just like the title says; can I *safely* cancel a partition resize operation while it is in progress? Here is what happened. I was trying to make room on a 1.5 TB HD to load an OS on to, to diagnose some unrelated problems. So I was going to make a 40 GB partition on it. The HD is almost full. There is about 100 GB free space, out of the ~1.3 TB in total. I resized the partition and made 40 GB of unallocated space before the existing partition. While doing this I thought to myself that this is dangerous and stupid, but I did it anyway because I really need to get something else fixed. As soon as I released the mouse click on commit, I realize this is going to take a long time, and started to wish I hadn't done it. But now, about 60 minutes later, the status bar STILL shows no progress for operation progress and total progress, and still shows no time left. So my my question is: can I *safely* cancel this operation while it is trying (apparently) to perform it? What I mean by safely, if it isn't obvious, is that I have about 1.2 TB of important data on this hard drive, not backed up, and don't want to lose it. I have been in a similar situation before, and remember that when I tried to cancel, nothing happened and it just kept on working on the resize. Thanks!
That's kind of what I expected, but I also expected it to show some kind of progress by now, or would have calculated an estimated time. It still shows no bars on either progress meter.
Theoretically, if I did try to cancel and data loss did result, how easy is it to recover? Is there any software that has good success rates with recovering data lost from this situation?
How great is the risk by merely hitting the cancel button? (as long as I let it run it's course if nothing happens when I click it)
I would be more assured if there was even a sliver of progress shown. What's your take if in the morning there still is no progress shown? This is a pretty powerful machine and I need it to work on some complex solid models that I'm getting behind on, which my other machine can't handle--even if it told me I had 50 hours left to go, that would be fine, as long as I have a number. (Unfortunately I can't run the software at the moment, because of the initial problem that got me into this mess)
Thanks
Is the hard drive activity light showing high or constant access? That would tell you if anything is actually being moved.
My rule #1 when making any partitioning changes is to make an Entire Disk Image backup of the drive before you begin (unless the data is not important).
Do you have a backup image of the drive?
If there is drive activity, I would let it go for a while longer and see what happens. If there is not any activity, then it's very likely that the process stalled or locked up at the beginning and may be completely unchanged or already corrupted.
If you cancel the procedure, the drive will be left in its current state. You won't know until afterwards what that state is.
That's kind of what I expected, but I also expected it to show some kind of progress by now, or would have calculated an estimated time. It still shows no bars on either progress meter.
Theoretically, if I did try to cancel and data loss did result, how easy is it to recover? Is there any software that has good success rates with recovering data lost from this situation?
How great is the risk by merely hitting the cancel button? (as long as I let it run it's course if nothing happens when I click it)
I would be more assured if there was even a sliver of progress shown. What's your take if in the morning there still is no progress shown? This is a pretty powerful machine and I need it to work on some complex solid models that I'm getting behind on, which my other machine can't handle--even if it told me I had 50 hours left to go, that would be fine, as long as I have a number. (Unfortunately I can't run the software at the moment, because of the initial problem that got me into this mess)
Thanks
Wow--that's really weird. 3 hours later when I was at the bar, I somehow made a duplicate reply.
As far as the drive activity goes, the HDD LED isn't connect at the moment, but I'm sure if I connected it in the morning, it should function properly. It is disconnected, because it used to be connected to my LSI RAID card, but my system took some kind of static shock anomaly a month ago that apparently went through the case to a hard drive, which knocked out my LSI card and wiped out my OS drive (raid 0). I think this is the last time I use a SECC case instead of aluminum. Of course I had a back up of this (ATI), but I haven't been able to get any backups to restore properly. Hence, installing a temp OS on a different drive to help diagnose. It's so stupid, because I forgot I had two unused 15K SAS drives in my desk drawer that I could have used instead. And I remembered them soon after I started the operation, which was slightly after I regretted doing it.
I digress--it's been a long week.
No, there is no backup copy of this data. My primary storage is a 1 TB RAID 1 array that gets backed up on a third 1 TB drive which is kept in storage. This 1.5 TB drive is overflow because the 1 TB array is usually full. It doesn't have critical data, but the data is still pretty important. Important enough that I don't want to lose it, but unfortunately I never purchased a second 1.5 TB drive to back it up due to it's temporary nature. You never figure you'll have this problem with a drive that isn't powered 24/7. (no it's not USB--it's eSATA)
So how easy is data to recover in this situation? And if there is no HDD activity, is it best to cancel it sooner than later?
I'm sorry to throw this at you guys. I'm still kicking myself for this-I have no idea how I let myself get into this-I knew better. I'm so stressed out at the moment due to non-related factors, that I think I let my guard down.
Thanks
DD should only need to move the 40GB of data, so you should have seen something. I suspect that it's not doing anything and may not have done anything.
Is the eSATA drive connected to a real eSATA port or is it a SATA conversion connection. I gave up on the conversion type because it wouldn't keep the connection to the drive.
If there's still no drive activity, then I don't think it will matter if you stop it now or later. Do you know for sure that DD is still responsive? DD may be locked up, in which case you may need to reset the computer.
I didn't hook up the HDD LED, but I forgot there is an activity LED on the enclosure, and there is no drive activity shown.
It's a bit of a conversion process. This is an internal SATA drive that inside an aluminum eSATA enclosure, so I suspect it gets converted right there.
As far as I know, DD is still responsive. I can move the progress window around--but that may not mean it is still responsive.
If the partition is corrupted and data loss does result, can you recommend any software/methods for recovering it?
If you can move the DD window around, it's probably not locked up and will probably cancel.
Regarding the eSATA connection, yours sounds normal. What I meant was not the connection in the enclosure, but the connection on the computer side. Some people use a standard SATA port on the motherboard and connect an adaptor to it. The adaptor sits in a normal slot opening and contains an eSATA port. This type of connection (in my experience) is not nearly as stable as a real eSATA port on the motherboard or on an add-in eSATA card.
I don't have a data recovery program to recommend personally. However, quite a few people have posted the use of GetDataBack. Most recovery programs have free trial versions so you can see if they find what you want before you need to buy them. I hope you don't need one.
So what do you think? Does it seem like the partition is unchanged and will cancel without problems?
I guess what I should be asking, is, it seems that at this point it is definitely stalled and not working on the operation. So my only choice, really, is to cancel it?
Are you running DD in Windows or from the DD CD? If in Windows, do you still have access to the drive (you may have if DD didn't get it locked).
It's really just a guess as to the state of the partition if you cancel. However, since it didn't make any progress there's a chance that nothing happened and no changes were made.
I'm running DD in windows. The drive is locked out. It appears in the list, but with no total size or free space shown.
I'm getting closer to canceling it. I clicked cancel and it asked me to confirm it, so at least that means it isn't likely locked up. I haven't chosen yes or no yet.
The last time I was in this position, when I clicked cancel nothing happened, but it was still working and progressing, just that 3 or maybe 8 hours into resizing a 250 GB partition on a 300 GB HD, it still wanted another 20 hours and somehow the laptop that this was run from went on standby eventually, which resulted in data loss. But somewhere in there it looked like it locked up, and I tried to cancel, and it completely ignored that action. (and eventually showed more progress, but never acknowledged the cancel request)
So I think that shows a little hope in this situation...
Ok. I clicked cancel, and clicked yes. But nothing happened. So I tried that again. Clicked cancel and confirmed yes, and I'm still at the operation progress window that shows no progress.
What do you think?
After trying to cancel, I'm left with the same progress window:
You probably need to terminate DD and/or restart Windows. I don't think DD is going to respond and exit on its own.
In the other case you mention, I would assume there was drive activity. There is nothing in this case.
Well, I'm much more weary about just restarting windows to stop it. Would terminating DD (I assume through task manager) raise the chance of data loss? Is it worth letting it sit for awhile after trying to cancel? Or if it doesn't respond to canceling now, is it not likely to ever respond? (since there is no drive activity)
To mention it for the record, I guess. In the other case, there was drive activity. For the most part, it was actively progressing. When the system came off of stand by, DD attemped to resume where it left off, and said it couldn't find such and such sector, and aborted itself. Which then resulted in a corrupted partition with data loss.
When you restart Windows, running programs will be terminated. You can also terminate them from the Task Manager (in most cases).
I doubt that anything you do at this point is going to increase the chance of data loss. Data may or may not have ocurred. I think DD got "stuck" at some point very early in the resizing process. If it happened before any changes were made, there's a good chance that there won't be any data loss.
You can certainly wait longer and see if DD responds and cancels out if you want. I doubt it will ever respond, but I have seen it take some time. However, in those cases, DD was working and showing progress and not in a "locked-up" state.
In the Task Manager, does DD show that it's using any CPU resources?
Oh yeah, good idea. Task manager shows DiskDirector.exe is using 13% CPU and 63 MB ram.
There is also 16 svchost.exe processes, which seems to be many more than normal. I'm just wondering if that is an indication that something else wasn't running correctly, and could have contributed to the problem.
At 70 currently running processes, that's far more than normal.
I'm running a few programs right now (10) and Vista is showing 81 processes running (including 14 svchost.exe processes). I don't think that has anything to do with the problem.
The CPU usage could be DD trying to do something, being stuck in a loop or trying to cancel. If you're going to wait a while, keep an eye on it and see if it changes.
What CPU do you have? 13% of an older single-core CPU is a lot less than if you're running a quad. Just keep in mind that whatever DD is doing, with no drive activity, it's probably not doing anything functional.
It's a 3.2 Ghz Core i7 (quad core), 12 GB ram. The average CPU time is changing, but the CPU value is constant at 13. It pretty much seems to be in the same state as when I first tried to cancel an hour and a half ago.
So, should I terminate it by ending the task in task manager?
I just canceled the process using task manager, and the drive was instantly unlocked in my computer. The data *appears* to be completely intact. At least the drive is in the same state as before I ran DD, and I was able to browse the folders.
I can't believe I just wasted 24 hours on this problem, but in the end, that's better than dealing with data recovery and probable data loss.
Thanks for your help and your time.
I'm glad the data appears to be intact. Hopefully, it really is. It would be a good idea to make a backup before trying anything further. A backup is recommended, in any case.
It's much, much better to have wasted some time and not lost any data. There have been cases of severe data loss because of stopping an in-process operation. I don't think yours ever got started.
You might want to run a chkdsk /f on the drive to check for any errors, but I would make a backup of any important files first.
I've worked with the data a bit now, and it seems pretty sound. It would appear you were right--I think DD locks out the drive as soon as you begin to mark changes, so after clicking commit, it probably didn't do anything beyond displaying the progress screen. I won't have to try anything further with the partition, because I was able to use that unused 15k drive I had in my desk instead. Much better than messing with something that huge just to troubleshoot some stuff for a week.
In regard to the running processes, I probably should have mentioned I'm running Server 2008. Even with the full desktop experience and aero enabled--right now I have ProE running, ansys, winamp, outlook, excel and firefox, and I'm only at 53 running processes. Granted, it's a bit relative to how much stuff you have running, but one of the nice things about WS08 is that it cuts out most of the bloat that vista has. I still have 16 svchost.exe processes running, (probably doesn't mean anything anyway) but if I had 70 running processes with only DD running, there may have been something wrong going on that interfered with DD...who knows.
Do you guys mark threads solved when they become so? I'm not sure where to direct this to, but in future versions, and with hard drives becoming so huge, DD should give a warning when you try to start an operation that will take more than, say, 24 hours, and results in data loss if you stop it before it finishes. Just a thought...
You can edit the title of the thread since you started it (use the Edit button at the top) and just add [SOLVED] to it. As far as I know, there isn't any built-in method.
The number of running processes totally depends on what is running. I just go by what I've seen in working on computers. I also get familiar with what mine normally run.
Hello all,
Nick,
I would like to express our gratitude to Mudcrab for the help is resolving the issue.
Nick,
DD should give a warning when you try to start an operation that will take more than, say, 24 hours, and results in data loss if you stop it before it finishes. Just a thought...
I will submit a request to the appropriate department. Thank you for the idea.
PS. I edited the title.
Thank you.