Nerva

OK, the BitFlock ID is W542ECTT.

Wait -- I installed the SSD plugin, and after adding the Seagate SMR drive to the pool, I wanted to set it to "Archive" and set the other HDD's in the pool to neither "SSD" nor "Archive", so that they are used for storage but are prioritized for new files rather than the Seagate SMR -- but the plugin will only let me toggle between SSD/Archive checkboxes -- I can't deselect both checkboxes at the same time. Could the plugin be updated to allow neither SSD/Archive to be checked and thus effectively create a third type -- a "normal" HDD?

OK, the problem has resurfaced. I rearranged which drives are connected to which controllers, and moved a 3TB drive onto the Supermicro AOC-SAT2-MV8 8-port controller -- and now all the 2TB and 1TB drives connected to the controller are reported as 3TB drives in Stablebit Scanner, even though Windows correctly identifies them as 1TB and 2TB drives.

Wait, so am I right in understanding that anything marked "SSD" will be kept entirely empty (only used for temporary writes)? So if I were to mark one of my 4TB HDD's as "SSD" and the new 8TB Archive HDD as "Archive", DrivePool will try to keep the 4TB drive empty, so it is "ready" to write new files? I do use duplication (indeed, the purpose of getting the 8TB SMR drives is to finally enable me to have 100% duplication), and I only have the one SSD (which will only have 60 GB or so free space), so am I right in thinking the thing to do is mark the 8TB SMR's as "archive", not bother to add the boot drive to the pool, and leave the non-SMR HDD's not marked either "SSD" or "Archive", so that DrivePool will only write to them, but then balance them equally with the SMR's?

My home server currently has a mix of 1TB, 2TB, 3TB, and 4TB drives, plus a 64 GB boot drive that is nearly full with Windows 7 crap, so I am about to swap it out for a 120 GB SSD, so there will be about 60 GB free on the boot drive. I've also ordered a Seagate 8TB Archive HDD, and plan to order one or two more, once I know they will work with my server. Of course, the issue with those drives is their goofy write behavior, and I have read that people recommend using the SSD plug-in to prioritize writing new files to the other drives, and then later the plug-in will move the files to the archive drive. What I'm wondering is, whether I should use the 60 GB free on the SSD as the landing-zone for new files, or should I just use all the other non-archive HDD's? I think this depends on details I don't understand regarding how the SSD plug-in actually works. If I use the SSD, I would want that 60 GB free to be kept empty and only used for temporary storage until the files can be moved. If I were to use the non-archive HDD's, I would want them to be about the same % full as the archive HDD's, write only to the non-archive HDD's and rebalance later. Can the SSD plugin be used either way, or just one?

Yes, everything looks fine now -- thanks!

I had 3 Western Digital 1TB drives and 3 Seagate 2TB drives connected to two Promise SATA300 TX4 controllers for many years with no problems. Just recently I replaced the pair of Promise controllers with a Supermicro AOC-SAT2-MV8 8-port controller, and Scanner is now mysteriously claiming that the 1TB WD drives are 2TB, even though the controller says they are 1TB during startup and Windows "My Computer" says they are 1TB -- at the same time, Scanner claims that two of the drives suddenly have 3.45 GB and 2.85 GB (respectively) in bad sectors that need to be tested. Scanner is also trying to automatically re-test all 6 drives connected to the new controller, but it is taking a VERY long time to scan the first 1TB drive that supposedly has bad sectors. Any idea what is going on??

Has anyone tried using Seagate 8TB Archive drives in their Drivepool setup? I have a home server with about 20TB of storage -- 90% of it is used for video (mostly my blu-ray collection) and the remainder is used for music and backups of computers in the house -- currently only a small amount of important stuff is redundant. I also use a couple old 1TB drives just for PVR that are in a separate pool. I'd like to add some larger drives (probably 8TB) so that I can make the entire thing fully redundant. I've been reading this review of the Seagate 8TB Archive drive that uses SMR: http://www.storagereview.com/seagate_archive_hdd_review_8tb The upshot is that read performance is excellent (probably 10x or 100x my bandwidth needs for watching movies), but due to the SMR technology the write performance drops from excellent to horrible if sustained writes are larger than 20GB. Basically it has 20GB of space it can write to quickly, and then it slowly moves that to the SMR portion of the drive -- so if you saturate the cache, it slows to a crawl. The article explains that this is horrible for a RAID array when it comes time to rebuild parity, but it is considered acceptable for non-RAID applications. So, I'm thinking that this drive should work acceptably with Drivepool, since the write operations are just to fill the drive in the first place and then to add new blu-ray files from time to time. The good thing about the drive is it has very low power consumption and price per TB, so it would seem to be a good choice for my home server application.

I didn't even notice those buttons -- looks like that fixed it.

I keep trying to figure out what I'm overlooking, because this seems like such a basic issue, it shouldn't be a bug, but I am finding that Scanner only rechecks the surface of my disks when I manually tell it to do so -- despite the fact my settings say they should be automatically rechecked every 30 days. I've waited for over four months and Scanner has yet to recheck a single disk's surface on its own -- the GUI says the disks were last scanned four months ago (when I manually started the scans) -- this is on a simple home server that is on 24/7 and has relatively low activity. I'm also concerned the same is true for the file system scans -- Scanner's GUI doesn't say when the last check was done, only that it has been done at some point. I have looked through the settings, and everything seems set up to facilitate automatic scans: 1) Scans can be performed any day at any time. 2) It throttles activity when overheating (it does not suspend activity). 3) It scans at background I/O priority, at medium thrashing sensitivity. 4) It doesn't interfere with other disks on the same controller. But even the little-used disks on a SATA3 controller aren't being scanned.

OK thanks -- I had it rescan the bad sectors and they are good. I'm having it do a complete rescan of the entire drive, just to be sure...

In my pool I had a 4TB drive that Scanner said has 2MB of damage, so I told DrivePool to remove the drive from the pool, but there was probably a hundred or so GB of files I was unable to read to get off the drive, so I told DrivePool to forcibly remove the drive from the pool, since those files could be recovered in other ways (with some effort). About six months later I discovered that drive had a bad SATA cable, and I quickly moved the remaining files off the drive (manually), so I didn't lose anything whatsoever, except the 2MB of damaged files, which were duplicated, so it wasn't an issue there either. Since I figured the SATA cable was probably the only problem all along, I removed all partitions from the drive and created a new partition with a full (not "quick") format, to ensure everything was good -- then I ran Burst Test for nearly a day, which it passed -- the the drive seemed good to go for being re-added to the pool. That's where things are confusing me. I added the drive to the pool, and I see there's a red arrow for "New un-duplicated file placement limit", which is zero, and I can't quite figure out why. Scanner still complains that the drive is damaged, so I disabled the option to remove files from damaged drives... but the red arrow is still at zero. I am wondering if DrivePool could be confused, since I had told it to remove the drive from the pool six months ago, but that was before I repartitioned, reformatted, and readded it to the pool.

Probably six months ago I got a notice from Scanner that one of my pooled (using DrivePool) 4TB drives was damaged (less than 2MB worth). I moved all the files off of it, deleted the partition (leaving the drive completely empty), and created a new partition with a full (not "quick") format. Scanner says the damage is still there, but (no surprise) there are no files allocated in the damaged area. However, Scanner loves to keep reminding me the drive is damaged, even though the damage has not increased, and I ran a Burst Test on the drive for an entire day, which it passed. My pool is 99% full without that "damaged" 4TB drive being included and that prevents me backing up the computers in the house, so I would like to add it back to the pool, since the damage area isn't in use. What is frustrating is that DrivePool won't put any files on it, because it is "damaged", even though the damage is old and quarantined. The "damaged" drive had yet to fail in six months, but one of my older 1TB drives has since failed and I lost some non-essential data because there was no room to duplicate it. I finally gave up and disabled the option to remove duplicated files from damaged drives, so that it would use the drive at least for duplication. I recently discovered that the 4TB drive had a bad SATA cable, so I am wondering if perhaps the "damage" isn't even real. However, I don't see an option in Scanner to manually re-check damaged sectors -- all I see is an option for how often it should automatically retest them. Anyways, I'm wondering if there's any way to recheck the drive, or somehow clear the "damaged" flag for old damage?

Thanks for the info -- I plan on building a new server "correctly" someday, but at the moment I am unemployed and so I can't afford anything besides hooking a new drive up to my current hardware. I may have to return the 8TB drive and just buy a 6TB, since those should work with the SYBA controller.

I contacted both SYBA and AMD regarding drivers -- SYBA was very responsive but they said they are at the mercy of getting drivers from Marvell, who makes their chipset -- and Marvell does not have end-user tech support to contact. AMD's response was terse, broken English that seemed to tell me to contact ASUS for a driver -- I shot back that all ASUS does is put their logo on AMD's driver installer -- I'm supposed to ask ASUS to ask AMD to update their chipset driver??

What controller(s) in your system do you hook your 8TB drives up to?

I run both StableBit Scanner and Drivepool on my home server. It already has two 4TB Seagate drives, but when I tried adding one of Seagate's new 8TB "archive" drives, I get nothing but problems initializing it. I've tried connecting to drive both directly to the ASUS E35M-1M Pro motherboard and also to the SYBA SYBA SI-PEX40064 controller that the 4TB drives use -- Windows Disk Management and Stablebit Scanner both report the former configuration is a 3.86 GB drive while the latter configuration is reported as 128 GB, instead of 8TB.

What does that matter? What does that matter? What does that matter? I've disabled the notification in the settings, so why am I getting it at all? It's a simple software question.

About a year ago I went into Settings > Notification Settings and disabled the notification for "Notify when a disk is overheating". Yet I've continued to see one popup after another notifying me of an overheating drive. Out of curiousity, I enabled email notifications, and now I'm being spammed with drive overheating messages. I'm currently running 2.5.2.3100 beta.

The real issue is why does Stablebit Scanner keep popping up bubble notifications for obsolete issues that I've already clicked on and opened the program to see the details of? The hard drive in question had been WIPED and REFORMATTED, and yet Scanner keeps telling me I have unreadable data on a disk with no data stored on it! And why does it keep giving me bubble notifications for issues that I've DISABLED the bubble notifications for??

I'm seeing what seems like buggy notification behavior from Stablebit Scanner. It started a year or so ago when I first installed it -- notification balloons would popup telling me that a given disk was overheating. The overheating was a known issue so I disabled the notification for overheating -- except it continues (for months now) to give me notification balloons about the same disks overheating, even though the notification should be disabled. I would also get notifications even when the drives in question were not above their temperature limit. Now, within the last couple weeks, it detected damage to one of the hard drives -- it said that 3.5kb are unreadable on one of the drives. It kept giving me the notification, so I took the drive out of service (removed it from DrivePool where it was being used to duplicate another drive). Yet it still kept giving me the "unreadable data" notification for the now-empty drive. So I decided to FORMAT the drive (not a quick format, a full format). And... it STILL tells me that 3.5kb of my data on a completely empty drive is unreadable. From what I can tell, the notifications are "stuck".

So, my server's SSD boot drive suddenly died. Fortunately I had a spare handy and was able to reinstall Windows 7 x64. I was overjoyed when after reinstalling DrivePool, my pools were immediately recreated as if nothing had happened. HOWEVER. It seems the files and folders have old security permissions left over, and they are interfering with using the files. Any time I try to move/delete/rename/edit any file from the server locally, it pops up with an administrator authorization, and if I try to do it over the network from my desktop machine, it says it is unsuccessful. So it seems like I need to cleanse the drive-pools of their old security permissions and reset them to their defaults. Is there an easy way to do this?

I'm debating between formatting my drives in 4k clusters (the largest you can use with NTFS compression) vs. 64k clusters (better for all my bluray movie files). But does the use of pooled drives complicate things? In particular, for 4k clusters there is a 16TB limit on volumes. I am curious if there are any issues with pools larger than 16TB if the individual drives use 4k clusters? Also, I am thinking of using NTFS compression for certain directories (like desktop computer backups) that have very compressible files that are never used -- but in order to do that, I would need to use 4k clusters, and this makes me wonder if there could be issues with compressing files on larger than 16TB pools. I'm guessing the right answer is to use 64k clusters regardless, but I would still like to get a better understanding of what the limitations are when you factor in using DrivePool.