nevergrownup

DrivePool can go for a while between official releases, but if you run the beta builds they are updated very often (daily or weekly is the norm). I've yet to encounter any significant bug from running the beta builds. You can also follow development status on the wiki page here: https://wiki.covecube.com/Development_Status

A fact which made me extremely wary of the newer Hitachi/HGST drives. WD was my least favorite until this run of luck on Seagate, but so far I haven't gotten burnt on the newer HGST/WD drives. As you say, all the manufacturers have had their share of issues. I just try to keep a mixture of drives and make sure to scatter my purchases across multiple manufacturing lots to hedge my bets. That and make liberal use of data duplication on my pool to save me when something goes sideways!

I've been burned too many times on Seagate drives to feel comfortable buying anything new from them without them being on the market long enough to get a good read on reliability. Seagate used to be by favorite drive manufacturer until 1TB+ drives became common. I bought five 1TB Seagate drives over the course of a couple of months only to have all five of them fail with a year and a half. All were RMA'd and half the refurbs died within six months so I stopped using the 1TB drives alltogether. I had a couple of 1.5TB Seagates that got bit by the firmware bug that required me to wire up a JTAG like tool to unbrick them long enough to recover my data. The 3TB drives I skipped because I heard quality control after the flooding was particularly shoddy. Then 4TB drives started appearing and the Seagate drives were dramatically cheaper than any other brand so I rolled the dice again and had a 4tb start throwing SMART errors within a month. A few months ago a coworker bought several 5tb Seagate externals so we could archive some backups at work. They work fine, except for some strange reason the last 500-750GB I write on any of the drives copies at USB 1.1 speed. Obviously, all of this is anecdotal and doesn't necessarily reflect everyone's experience with Seagate, but it's made me swear off of Seagate at least until I see them put out a generation or two of drives that are widely regarded as being solid and reliable. I just feel like their quality control is weak at best. I'm sticking with HGST. I have at least a dozen 1TB, 2TB, and 4TB HGST drives in service and the only failure I've had was one drive that overheated when a fan in my external enclosure died. Again, just anecodotal experience, but so far HGST has earned my trust.

Just FYI for those of us who tend to stay on the bleeding edge of the latest betas. I'm not sure if this is just me or if it will affect others installing build 630+, but 630 changes the default appearance of the pool drive to no longer appear as a removable drive. This caused Windows to assign my pool a new drive letter and apparently caused Drivepool to lose all of my file placement rules and balancing plugin settings. Fortunately it wasn't too hard to reconfigure, but just a caution that installing the latest beta might blow away your settings or confuse Windows a bit. Just the dangers of living on the edge. Hopefully this is an issue that was limited to me or that is easily corrected before the next full release.

Just updated to build 616. Working perfectly, thanks!

Awesome, I appreciate the follow-up! I'm on build 606 for now so I'm fully functional. Let me know if you need any logs or crash dumps.

I've narrowed down the build where the BSOD started. Everything works fine until build 607. As long as I stay on build 606 or earlier I can create symlinks and junctions. Update to 607 and it causes a SYSTEM_SERVICE_EXCEPTION every time.

Is anybody else getting a blue screen of death creating a symlink or junction on a pool in the latest betas? I'm running Server 2012 R2 and build 615 of DrivePool and I get a SYSTEM_SERVICE_EXCEPTION whenever I try to create a junction or symlink. I reverted back to the last official release (build 561) and it works correctly again. Not sure on what build the BSOD behavior started. Getting ready to keep trying newer versions until the the problem recurs. Just wondering if it's a problem isolated to me and my setup or if this is a more general problem.

Because of the nature of Hard Links I doubt that DrivePool will ever really support them. They are an NTFS filesystem construct to create multiple pointers in the filesystem index to the same file on a drive. Directory junctions and symbolic links operate on a different level. I'm sure someone can chime in with a more detailed explanation, but basically one is like street address that the post office (drivepool) can intercept and forward to a different address and the other is more like GPS coordinates that are fixed. That's a pretty weak comparison, but I think it hits the general idea.

I think the Ordered File Placement plugin could be used to do this. Setup your folder rule to place all music on drive 1 & 2. Then use the OFP plugin to designate drive 1 with a higher priority than drive 2.

Bingo, problem solved

Any idea why it's not preserving the rule order? Is there a log or something that would be useful in diagnosing?

This feature has been one I've been eagerly anticipating and I've got to say it is amazing. I love that no matter what happens to my pool I can take an individual drive and be confident that my entire music collection is there. No trying to merge the contents of five different drives to get my albums back in order if anything catostrophic happens. It's easy to set my documents on the fastest disk and set archival content on a slow archival disk. The sheer level of control is fantastic. This feature alone puts DrivePool head and shoulders above any competing product. So for all the effort to make this happen. Thank you. Thank you. Thank you.

This could be useful information. I've changed the the labels of my drives in Windows Explorer to show the size/primary contents of each just for my own reference (e.g. (Hitachi 2TB - Videos) E:) This would be helpful, especially for someone who doesn't have a firm idea of their drive's real world speed. Also lends itself to adding some sort of one click auto configure to setup a feeder/archive disk setup when you setup a pool. "We detected that one of your drives is significantly faster than the rest of your pool. Would you like it to be used as a high speed landing zone for your pool? Yes/No" I'm not sure how useful it would be to have it automatically retesting performance of a drive. As a rule the drive speed shouldn't change much. Maybe in the instance of an older SSD without TRIM you could use that information to zero out empty space on some sort of schedule? Not sure, just throwing out ideas. Yes, an Apply button would be nice. It's not clear at first that you can create a bunch of Folder rules at once. Normal instinct says if I click away from something and it disappears than it's gone forever unless I clicked Save or Apply. Yes, or at least some indication that file placement is being controlled by a parent rule unless you specifically override it.

Before changes. After changes, but before save. After saving and re-opening.