Christopher (Drashna)

Ah, if you're backing up, that would make sense. And yes, you can create a secondary pool, and copy the files to this pool. And once both (all) disks in that pool are offline, the pool would "disappear". That sounds exactly like what you want.

Honest, that depends on what you want. But running WHS in hyperV means that if your hardware fails, it makes it a lot easier to "port" over to a new system. But I think if you have everything setup already, that you should leave it that way. But the amount of RAM it gets, and the number of other running VMs can adversely affect the VM. As for the virtual hard drives, I would recommend passing the drives through to the VM. this will get better performance than hosting the VHD files on a HDD. Though, this is why a lot of people use a RAID5 array. Better speed and reliablity for virtual hard drives. But as I said, I think passing the disks through would be your best option. But personally, I feel that running it on the hardware directly will get better performance (that or upgrading the host to newer versions of Windows with HyperV).

Given what the error looks to be (related to Disk ID's), I suspect that the issue is definitely related.

otispresley, that's a very nice, higher end system. @Scar45: The Norco RPC 4220 is a very nice case. I know, I have one. As for SAS, for the most part, it's nothing to be afraid of. It's higher end storage stuff. It allows for higher speeds, and a lot more connections. But for the most part, that may not be much of a concern for you. And yes, the Norco has 5x mini SFF-8087 connects. As well as 5x power connectors for the drives. So you'd need those cables. Well, I'd recommend a "reverse breakout cable" to connect 4 of your motherboard ports to 4 of the bays. http://www.newegg.ca/Product/Product.aspx?Item=N82E16816133033 As for the controller cards, I have a HighPoint RocketRAID 2720SGL card. And "JBOD" isn't the term you're looking for (at least not with this card). If you do get this card, then you'd just want to attach the disks, and "be done with it". It will set the disks into "Legacy Mode", which is what you want. This will pass the disks to the OS, and will pass on the SMART data (if you enable "UnsafeDirectIo"). As for the other two, I'd recommend LSI over the other. But that's because I'm m ore familiar with LSI. I believe the BIOS for that allows you to just pass the disks through (but I'm not sure, as I hacked mine to use the "IT" firmware, which acts as just a controller card, which is what I wanted). However, with the cards that you have listed, you'd need two of them. And you'd need four of the SAS cables: http://www.newegg.ca/Product/Product.aspx?Item=N82E16816133034 (or 3 cards and 5 cables, if you don't get the reverse breakout cable and connect 4 of the motherboard ports) Either way, once you get the parts and the server installed, and the drives connected. As soon as DrivePool is installed, if it sees the drives, it will immediately recreate the pool. You will need to "reshare" the folders. To do that, download the WSS troubleshooter, and run the "Restore DRivePool shares (reinstall)" option: http://wiki.covecube.com/StableBit_DrivePool_Utilities Also, if you do buy any hardware, I would recommend boards that support ECC RAM. This is more expensive (both for the RAM and the mobo usually), but it's more reliable, and less likely to experience issues. That, or most AMD boards do support ECC RAM, but check.

Well, I'm glad to hear that (almost) everything went very smoothly! Though as for junctions, DrivePool v2.1.0.443 should actually support them, if you really do want to use them. As for the performance, you should see performance that is akin to a normal drive (as we're passing the operations directly to the disks). That and there are some performance tweaks too (the network IO boost). But I'm glad that you're enjoying better performance, be it from the new hardware or the new software. As for the externals, that's a tricky one. Ideally, the disks should be left on for DrivePool. The reason is, that any time disks go "missing" or are turned off, the Pool that they are a part of, well a) reports the issue, sets the pool into a read only mode. This is expected and normal. However, it may not be ideal for what you want. Though, you should not be losing your settings. Could you upload the ErrorReports for us? http://wiki.covecube.com/StableBit_DrivePool_2.x_Error_Reports However, if you're turning the disks off for power saving, StableBit Scanner *may* be able to help with that. But that really depends on the controller and the disk though.

To clarify here: After the first license of EITHER product, you can get 10 additional license of EITHER product (or any combination of the two, that adds up to 10) with a $9.95 discount on the price. The "Combo" item has this second license at a discounted price (and counts against the 10 limit). So if you already own DrivePool, and Scanner, you could purchase 9 additional licenses of Scanner. Just note, that all the licenses will use the same Activation ID.

First, I'm sorry to hear that one of your disks is dying. That's never a pleasant experience. As for the disk, there is a "Force removal" option, that will skip any errors and just remove the disk from the pool. This is what you want. Once that is done, it may leave a lot of your data on that disk. However, it should be accessible in a "PoolPart.xxxx" folder (that was hidden but is now exposed). You can copy these files directly into the pool (the contents of the folder will mirror the pool's folder and file structure/layout). If you want, we could do this remotely, for you, so that you feel safe with your data.

For duplciation: this affects new files mostly. However, when realtime duplication is enabled, it will also make passes periodically to ensure that the files are duplicated properly. If it is disabled, then it will only check once a day (2am by default). As for disk removalL When you remove a disk from the pool, the content is moved to other disks in the pool. Depending on the size of the disk and the amount of content on it, this can take hours. However, both the "Duplicate later" and "force removal of damaged disks" options will leave files on the pool. (and may make the process quicker).

I can definitely understand that.

Yes, that's exactly what it will do. However, I wouldn't say it won't "complain". When it makes a duplication pass, it will note the number of copies there are, and how many there should be and copy the files or delete duplicates, as needed. But yes, it will definitely "aggregate" the folder contents from both drives on the Pool.

Well, for the security issues, running the "Reset permissions on the pool" option in the WSS troubleshooter may help. http://wiki.covecube.com/StableBit_DrivePool_Utilities

There is also a "force remove" option, that will skip file errors, instead of stopping at them. This will allow you to remove a damaged disk that is having issues. However, it may leave files in the "PoolPart" folder (which will be exposed once it's removed from the pool). You may want to copy these manually, just in case. Alternatively, you could add this third disk to the pool, and manually move the contents of the failing disk to the other drive. (or just remove the disk afterwards and let DrivePool handle that).

This may not be exactly for your system, but it should give you an idea of exactly how to convert: http://wiki.covecube.com/StableBit_DrivePool_Q5463715 And yes, the contents will be merged when it shows the pool's contents. For folders, this is fine and won't pose any problems. However, for files: As for the pool, once it makes a duplication pass, if it detects the same files on multiple drives, it will check whether or not if it's supposed to be duplicated. If it's not, then it will delete a copy. If it is supposed to be duplicated, it will leave the copy. However, if the files are different, but share the same name, it will prompt you about that (pool parts different).

Taking ownership of the folders usually works, but yeah, it can be an issue dealing with. Also, what OS are you using, and what version of DrivePool? As for the disk dropouts, are the drives actually still present? Event viewer show any disk related errors? As for the "not enough disk space", that's an odd one. Could you enable tracing and duplicate the error? v1.3: http://wiki.covecube.com/StableBit_DrivePool_Log_Collection v:2.x: http://wiki.covecube.com/StableBit_DrivePool_2.x_Log_Collection

I would definitely recommend disabling duplication for these folders. Well, at least the "System Volume Information" folder. The reason is, that this folder is what contains all of the "shadow copies", which we don't support. No reason to waste additional space. As for the Recycle bin. that's up to you. However, if you're only accessing the files from the network, then it will not be used anyways. As for the pie chart, usually running the "Remeasure" option fixes most issues with it. As for why it didn't right away, do you have RealTime duplciation disabled? As for the Scanner process, could you get the Error Reports from Scanner? WHS: http://wiki.covecube.com/StableBit_Scanner_for_WSS_Error_Reports WIndows: http://wiki.covecube.com/StableBit_Scanner_for_Windows_Error_Reports

Okay, I just wanted to make sure then. And if that is the case, then feel free to do whatever you want with the D:\ drive. It shouldn't adversely affect the system.

Well, normally on WHS2011, it defaults the "Server Folders" to the D:\ drive. But I suspect that you've already moved the folders/shares off. If that is the case, then there should be no issues then.

Two works for you Ikon: Trouble Maker. Because I had to double check to see if this would even work first... (because I wasn't sure). Stop the DrivePool Service (and any service that may be using the files from the pool) Move the files out of the hidden "PoolPart" folders Delete said PoolPart folders. Start the DrivePool Service You'll see an empty DrivePool drive until you start the service backup. But I can confirm this definitely does work. However, it did not like messing with the "system volume information" folders in these poolPart folders. You probably will need to take ownership of the folder, and then delete it to get it to work/delete.

If it's the SMART error, then it has two options: ignore current, and ignore future. Pretty self-explanatory. But yes, if you choose the "ignore current" option, if it increases, it will warn you again.

Darkeuh, I know you don't have a "dashboard.exe" process, but you should definitely have a "DrivePool.Service.exe" process. The process for getting the dump is identical here, which is why I linked that article. It was originally for getting dumps of the WHS2011 dashboard, but as the process for any other .... well, process is the same, it's easier to reuse the same article. So basically, use that method to get create a dump of the "DrivePool.Service.exe" process, instead of the "Dashboard.exe" process.

The SYBA card you linked is a good one. I've used a similar chipset, and did not have any issues with it. Marvell makes pretty decent chipsets. And don't get me wrong, I'm not trying to rag on SYBA. The issue is with the Silicon Images chipset. I've been in the home server community since ... well WHSv1 was released (end of 2007), and I've seen a LOT of issues with there chipsets, and have experienced them myself. And Alex has had his share of issues with them, trying to get Scanner to read the SMART data from the drives, as I've said above. So, definitely avoid Silicon Image chipsets. As for the HighPoint cards, they usually use Marvell, so they're pretty decent. And if it's only for a 4 port card, then they should be good to use.

The reason we don't give the option to hide the drive is that the drive already exists, and we don't want to interfer with your system in a drastic way. Additionally, if there are any system files on it (like shares, or the client backup database), then changing or removing the drive letter would be detrimental to the system, and could cause issue (at least temporarily). And yes, you can manually change the drive letter, or even remove it. But we try to stay as "hands off" of your system as we can. But there is no issue with DrivePool and changing or removing Drive letters. Also, I would recommend mounting the drives to folder paths. The reason is, that you can still access the disks (and run maintainance, like chkdsk, on them easily, but it cleans up "My Computer", so you don't have a whole bunch of drives listed there: http://wiki.covecube.com/StableBit_DrivePool_Q4822624

darkeuh, That is not typical behavior, but unfortunately this does happen every once in a great while. Could you upload the error reports? http://wiki.covecube.com/StableBit_DrivePool_2.x_Error_Reports As well, as get a memory dump of the "DrivePool.Service.exe" process? http://wiki.covecube.com/StableBit_DrivePool_Dashboard_Freeze Once you've done those, reset the Settings for DrivePool: http://wiki.covecube.com/StableBit_DrivePool_Q2299585B After doing this, you should have to activate again. But this should be the last time that you have to do this. Regards

Eagleknight... I'm very sorry for your whole ordeal. That's not pleasant. As for the Client Backup Database, we highly recommend duplicating it. Because all it takes is one damaged file, or even one missing file to destroy the entire database. Trust me, I've been there and done that. A restore can fix the issues... but it usually does so by salvaging records. Which means you will lose backups. And while having crashplan is nice.... it's nowhere near as fast for recovery. As for your card.... when you describe what was happening with it, my first thought was "I really hope it isn't an Silicon Images chipset". And yup, it is. Sil3124 specifically. I hate these cards. So does Alex. Not only do that cause issues with Scanner (it only properly reports the first disk on the card, for SMART data), but they tend to fail under high loads. Which is precisely what you were seeing. And that's not limited to AMD or Intel. It's in any system. These chipsets are basically garbage, and should be avoided. If you want, I can give you suggestions for cards to replace them with. But I would recommend just about anything that uses Marvell. And if it's open, one of the PCI Express ports, as you'll get better speed that way.

Could you do this? http://wiki.covecube.com/StableBit_DrivePool_Log_Collection Don't enable the tracing option, just upload the folder. Use the thread's URL in the comments in the widget at the bottom. We can take a look and see if it's having issues. However, it shouldn't take 3 days to duplicate 500GBs (or less) of data. But that may depend on how many other disks are in the pool.