InlineRanger

I may have hit another snag... I can access the webgui log-in on a Win 7 machine using http://tower, but I can't on OS X using tower.local. However, I can access the webgui log-in using the IP address. Am I missing something about how OS X/Safari caches credentials? Everything in Safari Preferences -> Passwords related to UnRAID has been deleted. EDIT: Search yielded another user with the same issue. And another.

Wow! thanks for the help everyone. I never would have guessed that...

After a fresh re-install, I set my sole 250GB disk as "disk 1," made sure "user shares" was enabled under global settings and started the array. Changing the root password again caused the webgui to freeze. I captured the log and included it as an attachment. The server seems to be fully functional outside of changing the root password, but I don't want to commit data to it until it's working 100%. Thinking an SMB conflict might be an issue, I did another fresh re-install and tried disabling SMB and enabling NFS in the settings menu. Changing the root password causes the text "shutting down SMB... starting NFS..." to display at the bottom before the webgui freezes. Could this possibly be caused by a loose connection, incorrect BIOS setting or something equally drastic (zapped circuits?)? I doubt it's a kernel/UnRAID issue since the gen. 7 microservers are somewhat popular. Thanks for the help. I'm running out of ideas... tower-diagnostics-20160508-0251.zip

The log I posted was with my drive formatted using OS X's disk utility. The only relevant format option was MS-DOS (FAT), as shown in Lime-Tech's video. The video used an older version of disk utility and didn't show which "scheme" to use, but based on searching the forums I used "Master Boot Record." To double-check, I found a laptop running Windows 7 and reformatted the drive by setting the file system to "FAT32 (Default)," allocation unit size to "4096 bytes," volume label as "UNRAID" and left "Quick Format" checked. I extracted UnRAID 6.1.9 to the flash drive, ran the make_bootable batch file as administrator and still ran into the same webgui freeze problem. I captured the log again. To be thorough, I reformatted with "Quick Format" unchecked. Then I tried with another USB stick. The webgui problem persisted. What's so confusing is that I've gotten UnRAID 6.1.9 running on this machine when I last tested it a few months ago... EDIT: I do have an Airport Time Capsule serving files via SMB on the network. I wonder if that's causing a conflict? tower-diagnostics-20160507-2316.zip

My server is a bone stock HP N40L (2GB RAM, 250GB HDD formatted as XFS) running the trial version of UnRAID 6.1.9. My laptop is a Macbook Air running OS X El Capitan. I've been following the UnRAID "Get Started" video series and have successfully accessed the webgui and installed the temporary license. When I try to change the root password, the text "Restarting SMB..." appears at the bottom and the webgui freezes. The server still responds via ssh (requiring the updated password!), so I captured the diagnostics and powered down the server. Everything else seems to work fine if I skip changing the root password. I also tried UnRAID 6.1.8 and 6.2.0 beta 21, but both exhibited the same behavior. Any ideas? tower-diagnostics-20160507-2048.zip

If you're only storing a few Terabytes of data, I'd buy a 2-bay NAS and two 8TB disks -- or wait awhile and buy two 10TB disks. Depending on your workload, you could even buy two 8TB external drives, hook one up to your router and periodically sync it's contents to the other drive kept in cold storage.

Do any 6-platter drives have center mounting holes? I know the 8TB Seagate SMR drives don't. I think 7-platter drives use the same internal z-height, but with thinner platters. This was the article I remember seeing and it looks like storagereview.com corroborates. That would be great news if it's a 6-platter drive. The only 8TB PMR drive I know of is Seagate's enterprise ($$$) unit. The sooner mainstream 8TB drives transition to PMR the sooner we'll get 10TB SMR drives. I'd be surprised though, since WDC usually trails Seagate in platter density.

I don't have the source at hand, but I remember reading that they are essentially repackaged HGST He8 drives. As a 7 platter drive, these have slightly lower sequential throughput than the 8TB Seagate SMR drives (178MB/s vs 190MB/s max sustained read), but obviously don't have the random write slowdown for huge transfers. They also have a slight power consumption advantage (-2.5W spin-up, -1W operating, -0.25W sleep).

It looks like they use a proprietary form factor. I think they're too narrow to fit in a 5.25 bay and they lack mounting holes. They would probably work if you can figure out a way to jury rig them into your case.

Technically he can own it in 95 years when the copyright expires. We've waited over a half-century after Walt's death, but it's almost here! Steamboat Willie will finally enter the public domain in 2023.Just kidding!

Some food for thought is that 10TB SMR drives might be just around the corner. Seagate recently released 2TB SMR laptop drives using the requisite newer, higher density platter technology. There are a few reasons why it might be worthwhile to wait. The first is the obvious limitation that your parity drive has to be the biggest so go as big as you can. The second is that PMR drive technology is going to hit a wall around 12TB, IIRC. The move to HAMR may take longer than expected, meaning we could be stuck at 10-12TB for longer than usual, and when they are released their reliability will be unknown. Lastly, 10TB is a nice, round landmark number. Take part in history!

Interesting! I'm in the final stages of building my server, so I haven't really used TVHeadend yet other than trying it out on my Raspberry Pi. Thanks for the suggestion. My goal is to create a "sneakernet" separate from my internet-connected LAN, primarily for philosophical reasons. If my motives were entirely logical, I'd probably stop using UnRAID and just subscribe to Netflix/Hulu/Amazon Prime like everyone else.

With a UPS and a lot of RAM, my need for a cache drive has disappeared. I too would prefer an officially supported apps drive for the reasons RobJ mentioned.

Last time I hung around these forums, people were grumbling about v5.0 taking forever. Has that been released yet?

I have zero files in my cache directory/disk. The MediaTreeCheck shows the following: Media Tree: Detected DVB Adapter: Not Detected I'm pretty sure you're using the firmware for the HVR-2250, which is the old model that recently went EOL. Linux drivers for your card, the follow-on HVR-2255, were published last month. You can download them from the Hauppauge website. They are also included in the 4.2 Linux kernel. Do I need to ask the OpenELEC maintainer to include the drivers in his next patch? It looks like I'm trying to do the opposite of you -- move away from HDHomerun to a PCIe card. HDHomerun tuners are extremely reliable and serve streams over HTTP, but they're another box, ethernet cable and wall wart to deal with. One of the major appeals of UnRAID for me is (stably!) consolidating functionality into one box, a process I'd like to carry as far as possible. Kodi 16 and TVHeadend 4.2 have made a lot of progress, especially with respect to time-shifting. Things were buggy enough in the recent past that TVHeadend was moved to Kodi's unofficial app repository. AFIAK, there is still one major long-standing bug where playing back an in-progress recording won't update the recording size, causing the user the get kicked out to the home menu when playback reaches live TV. Otherwise, I think it may be ready for primetime as a DVR.

NTP servers require internet access. I'd like to run my server offline. Like you say, it's relatively easy to use UnRAID as an NTP server to keep other computers on the LAN synchronized. I've done it in the past. The tricky bit is maintaining accurate time on UnRAID offline. Fair enough, I understand this is a niche use case. If setting time from an ATSC signal is anything like using a GPS receiver, I don't think compiling a special version of UnRAID is even necessary. My goal was to open up a dialog on the subject, here and on more general purpose Linux forums, to help frame the problem. If I make any progress I'll document my findings to help anyone trying to go down this path in the future. Thanks for the input, everyone.

TVHeadend now includes an ATSC EPG parser, making it much easier to run a full-fledged media server offline. One of the last major roadblocks is getting (and maintaining) accurate time. UnRAID uses the system clock by default when it can't access NTP servers. While it works and is accurate, most system clocks drift too quickly for accurate PVR duties. Having to manually reset the clock is also a hassle. Many people maintain accurate time using GPS receivers, but that requires line-of-sight access to the sky, which many people may not have in their homes. Fortunately, accurate (enough...) time is broadcast as part of the ATSC signal. Many PVRs capable of operating offline (e.g. Channel Master DVR+) use this method. This is a niche request, but I think it opens up an interesting new use case for UnRAID.

I agree 100% and appreciate your effort in creating this guide, but I'd like to add my perspective as a Linux novice of why I plan on going with UnRAID despite it's closed nature. The content of my server vastly exceeds the cost of an UnRAID license and hardware, if nothing else than in time spent. Your solution appears more complex, which eats time and reduces stability. The fear of losing a lot of data -- quickly -- is too high. If someone packaged a minimal Debian install, MergerFS and Docker into a small, flash-drive sized package sporting a simple GUI, I'd switch in a heartbeat... even after paying for an UnRAID license. I personally have no need for virtual machines or real-time parity since my server is dedicated towards media storage. Once HDD's get big enough, I'll switch to a two drive RAID-1 setup because complexity sucks. I'm not a latest and greatest sort of person. Office programs and OS's largely stopped innovating over a decade ago, but the hamster wheel is still turning. The UnRAID wheel is worth running on at the moment because despite the fear of a flash drive dying and Tom going AWOL, it gets the job done with the least amount of complexity.

Wow, great work. Parity check speeds definitely look like they're CPU constrained. Too bad my N40L has about 65% of the processing power of the N54L, or ~750MB/s of total bandwidth if performance scales linearly. Looks like I'm stuck at A-Link speeds. Again, thanks for your help.

It's useful if you want to rip multiple discs in a row and not babysit the process. The issue would be minor if you have a fast drive and all discs rip consistently quick, but that's usually not the case. Some people are stuck with slow external/laptop drives, some discs have mastering errors/damage, some discs are SL while others are DL, and sometimes the drive just rips at half speed because... who knows why. A notification system can save a lot of time and mental accounting. Put disc in, start rip, do something else, disc pops out, "Oh, rip's done," put new disc in, start rip... ad infinitum. If you aren't interested then no problem. It's just an idea I thought might be useful.

Would it be possible to add auto-eject via a BASH script or similar? It's one of those MakeMKV requests that has been on the back-burner for years.

That makes sense. My idea of limiting the music share to one disk was to prevent more from spinning up every time I enter the music library. I'll play around with settings as you suggest and see what works best.

I'm really looking forward to your results. An interesting test might be using your 1430SA and splitting 6 disks evenly between the motherboard and controller card for ~250MB/s of potential bandwidth per port. Garycase's suggestions might also be useful.

My server will primarily store media (i.e. movies, tv shows, music), so a large portion of those million files would be in the music share. I was thinking about assigning my music share to a single drive and then setting cache_dirs to only scan the movies/tv shows shares. Browsing the movies/tv shows library would cause one disk to spin-up on playback, whereas browsing the music library would cause one disk to spin up immediately. Do you think that is a reasonable strategy?

Dual channel helps parity checks on server with many disks (12+), if I’m remembering correctly got an increase of about 10 to 15% on an Intel server and a little more on AMD. Do you think it depends on disk number or total bandwidth? In other words, would 1x 8TB drive see the same benefit as 2x 4TB drives? Dual channel seem like it would help my anemic CPU deal with the I/O of parity operations, but you're the first evidence I've seen. If you remember from the other thread, I'm the guy hoping to jam 16TB drives into an N40L.