Christopher (Drashna)

Well, everyone has different amounts of storage as per their needs.

Well, also, Emby (MediaBrowser Server) doesn't use hardlinks for it's metadata, so so you can place it in the pool without any issues. And it seems to use about a third of the space for metadata (plex uses ~150GBs, while emby uses ~60GBs)

Well, don't let me push you into anything. I just wanted to let you know about my experience and opinions with a similar circumstance. And, well, enjoy! And sorry to hear about that RAID5 array, I'm sure that was a "fun" day.

Sorry, usually the File History Backup feature backs up Users folder. Especially as this is the default location for these folders. In this case, using "\\localhost\f$\SomeFolder" is what you'd want to do.

I've checked for a few of the major changes that we made: 2.2.0.564 (antivirus related) 2.2.0.566 (BSOD related, "spin lock, but 32-bit systems only) 2.2.0.586 (Folder delete issue) 2.2.0.598 (BSOD related, concurrent directory listing issue) 2.2.0.624 (low level priority stuff) Needless to say, try the builds before these ones and see which triggers the issue

Well, thanks! And yeah, see if the firmware fixes the issue first. And jumping by 10 versions should ... be more efficient.

Okay, thanks. So, basically, not far enough back to make it easier to determine what change caused the issue. If you are a glutten for punishment, you can find all the beta builds here: http://dl.covecube.com/DrivePoolWindows/beta/download/ However, there is absolutely no obligation to test the builds here, at all. I only wanted to post the link, if you really want to. But if you do so, and are able to pinpoint a specific build where this starts happening, then that will greatly help us out. Again, ABSOLUTELY NO OBLIGATION to test these out. So don't feel pressured to.

Well, with how much they've been dropping in price of late.... I'm looking at getting a couple for myself, just to optimize my pool.

"Soon". We are waiting on the new version of Visual Studio to be released, so that we can properly compile a native Windows 10 version. Once this is done, we will compile it, test it and release it as soon as possible. In fact, we hope to do so before the July 29th release date, but no guarantees. So we're waiting on Microsoft.

Quality wise, Emby is the winner, clearly. And the UI? So much better, by far. Especially if you support them and get access to the CoverArt plugin. In fact, I've had friends compare it Netflix, because of the UI. Plex ... isn't there. But as for streams... everything I have is 720p or higher. And when transcoding, I can get 1, maybe 2 streams before it starts stuttering. But I've seen people that have some very nice systems (dual CPU Xeon's) that can transcode 20 streams without an issue. I'd also love if Emby would compile a version of ffmpeg that used the GPU accelleration. I got 4-6 times the performance out of it... but it was unstable. FFMPEG would fail to close properly, and block other streams from starting. And that is if it started.... But 20-30 fps vs 130-150fps .... a HUGE difference. And that's with just the onboard video for the Xeon E3-1245v3 CPU!

It's a well discussed (as in, I've mentioned it repeatedly to Alex, the developer) feature, that we do plan on adding in the future. The best option I can offer for now is to enable QoS on your router, to help curb the usage.

To be blunt, I cringe every time somebody says this. I was the same way... especially with the fact that I have about 20TBs of media (it's probably closer to 25TBs right now). I lost a drive, and lost a huge chunk of my media. It was a royal pain to recover all of it. This is even using apps like SickBeard and Plex to manage and list missing files. Some of the stuff you can only find in torrents ... that are poorly seeded (if at all). Some of it, I had to get from friends. Some of it ... I just deleted because I couldn't find the missing files. Yes, it is simple to recover ... but having the data duplicated means that all you have to do is remove the missing disk, and DrivePool will rebuild the duplication for you. No mess, no fuss. As for the controller card, if you ever need more drives, look into "SAS" cards. HighPoint has a few nice cards (though expensive). Otherwise, LSI cards are really, really good. Also, you can find these cards for cheap on ebay and the like, as large companies will sell them by the 100's or more, to recoup upgrade costs. And I'm glad to hear that you're already very happy with our products and willing to recommend them to others!

Thanks for the clarification. There have been a bunch of low level changes since the 2.1.1.561 release, so ... it's possible that it's triggering the issue. Also, what 2.2 beta build are you using specifically?

Could you check the balancing settings? Set it to balance immediately, but make sure the "not more than..." option is unchecked. Set the ratio slider to 100%. And see if that helps. If not enable logging and let it (or force it to) balance. http://wiki.covecube.com/StableBit_DrivePool_2.x_Log_Collection

For anyone wishing to use OneDrive for Business, we've fixed the authentication issues. Please download the following builds, if you're using this provider: 32-bit: http://dl.covecube.com/CloudDriveWindows/beta/download/StableBit.CloudDrive_1.0.0.296_x86_BETA.exe 64-bit: http://dl.covecube.com/CloudDriveWindows/beta/download/StableBit.CloudDrive_1.0.0.296_x64_BETA.exe

Glad to hear it! And if you have any other questions, don't hesitate to ask!

Ah, okay, there were in the "wrong" folder on our server. That's why I was having issues finding them. Sorry. And from the sounds of it, it does sound to be very related to DrivePool (or at least triggered by). I've flagged the dumps for Alex, and he'll take a look at them. But they're definitely related to the 3ware card, from the looks of them.

j That's kind of weird that it's happened, actually. But yes, you can safely remove the drive letter from that volume.

Well, thank you for posting the event viewer logs. From the looks of it, the 3Ware driver was having some issues right before the bugcheck (BSOD) occured. If the firmware update helps, then that's great! Otherwise, please do upload the crash dump (C:\Windows\MEMORY.DMP), and maybe the minidumps (the contents of C:\Windows\Minidump), so we can take a look at what is going on. It may give us a clue as to what is happening exactly.

Yes, and no. Specifically, you need as number of SSD drives equal to the level of duplication you're going to be using. This is because the Real Time Duplication feature writes to all the necessary disks to ensure that they are properly duplicated. If you disable realtime duplication, it will bypass this issue, but at a cost. First, files won't be duplicated until a duplication pass occurs (nightly, I believe). That means new files may not be properly protected. Secondly, files that are kept open (locked) may not be properly duplicated until they're closed. For these reasons, we recommend against disabling real time duplication, and grabbing a second SSD.

For anyone else having this issue, this is a weird issue with VSS, actually. Run "vssadmin add shadowstorage /for=c: /on=c: /maxsize=5GB" and reboot the system. This should resolve the issue.

That's.... really odd. There are three places I'd look here... Run a memory test. Use whatever tool, but run an extended test, and let it run overnight. Run a CPU test, use whatever benchmarking tool or Prime95, or whatever. Check the system drive. Run StableBit Scanner, and do a burst test overnight. See if it errors out. Or run the manufacturer's tool to check the system drive. Also, check the Event Viewer. See if there is anything that happens specifically, right before this happens. And check the Task Scheduler... maybe it's kicking off a task that is triggering this.

HyperV still doesn't support USB passthrough, sadly. And even Gen2 VMs don't really let you change the hardware either (it doesn't let you change the RAM or CPU cores, which is what I was talking about, but in both, you can always add and remove SCSI drives without powering off the system). There are ways to do the USB passthrough, but they're commercial software. Though, USB passthrough is really an edge case. Depending on what you need it for... it may not even be necessary.

Well, it sounds like you ahve a nice setup. One major difference between ESX and HyperV, ESX lets you change hardware on the fly. That may not be a huge thing for you, but it may. However, HyperV works really, really well in a domain environment. Really well.

That's definitely odd. I'm assuming that the boot disk was device 0 on the SCSI adapter? As for the "bootmgfw.efi", this is the UEFI boot loader for Windows. It's like "bootmgr" file for BIOS systems. The only thing I can think of is that the boot order was getting screwed up when adding the other drives. And that is very weird, as I use a bunch of Gen2 VMs in HyperV (Server 2012R2), and have never seen that issue.