So one of my drives was having errors and I went through the process of replacing the drive. Now that the data rebuild is complete, I am unable to view any files on the server. I pull up the directory in W7 but it only shows "This folder is empty." I can still play movies off the server from my media player, but cannot view anything from any computers. Any ideas?

 

If it matters at all, I've been having several disk issues w/ false errors recently. I was in the process of a data rebuild on one disk(17) when the disk mentioned above(1) gave errors. Once Disk 17 finished rebuilding, I shut down the system, replaced Disk 1, and started rebuilding that. I know I was able to view server files while 17 was rebuilding, and I believe I was also able to when 1 started rebuilding.

What does your webgui main page show?

Everything looks in order on the Main tab. All drives are there and the size and free space all looks correct. I can also view the contents of each disk by clicking on the folders.

It looks more like a network issue. Try a reboot of the server and maybe your router as well.

Run the New Permissions utility.

Might have spoke too soon about the webGUI. I played around with it a little when I got home from work and under the "Shares" tab, the media folder is there. When I click on it though, it's completely empty. The disks under "Main" still show their contents, but when I clicked Disk 1(just rebuilt, see above), it froze the system. I lost all network connectivity and the parity check stopped completely (was at 46%). I restarted the system and now it is rebuilding disk 17 again. Any ideas on what could be going on? I know system logs help a lot, but I'm not exactly sure what would be needed to get more detail. I also could not run the New Permissions utility since the parity check was running.

 

I had this system working perfectly for months until I attempted to expand using the Sans Digital 4 disk bay. I'm really tempted to replace that and see if that helps with any of these errors. Can I just replace the bay with another brand's and swap the disks out, or is that not plug n play?

I've rebooted the server, the router, and the computers I'm using to browse the server and still can't view the shared folder. When I click on Disk 1 using the webgui, I get the following error:

 

"Warning: filetype(): Lstat failed for /usr/local/emhttp/mnt/disk1/Media/.DS_Store in /usr/local/emhttp/plugins/indexer/Browse.php on line 21 Warning: filectime(): stat failed for /usr/local/emhttp/mnt/disk1/Media/.DS_Store in /usr/local/emhttp/plugins/indexer/Browse.php on line 31 Warning: filetype(): Lstat failed for /usr/local/emhttp/mnt/disk1/Media/._.DS_Store in /usr/local/emhttp/plugins/indexer/Browse.php on line 21 Warning: filectime(): stat failed for /usr/local/emhttp/mnt/disk1/Media/._.DS_Store in /usr/local/emhttp/plugins/indexer/Browse.php on line 31 Warning: filetype(): Lstat failed for /usr/local/emhttp/mnt/disk1/Media/Staging in /usr/local/emhttp/plugins/indexer/Browse.php on line 21 Warning: filectime(): stat failed for /usr/local/emhttp/mnt/disk1/Media/Staging in /usr/local/emhttp/plugins/indexer/Browse.php on line 31 Warning: filetype(): Lstat failed for /usr/local/emhttp/mnt/disk1/Media/TS in /usr/local/emhttp/plugins/indexer/Browse.php on line 21 Warning: filectime(): stat failed for /usr/local/emhttp/mnt/disk1/Media/TS in /usr/local/emhttp/plugins/indexer/Browse.php on line 31 Warning: filetype(): Lstat failed for /usr/local/emhttp/mnt/disk1/Media/TV Shows in /usr/local/emhttp/plugins/indexer/Browse.php on line 21 Warning: filectime(): stat failed for /usr/local/emhttp/mnt/disk1/Media/TV Shows in /usr/local/emhttp/plugins/indexer/Browse.php on line 31"

 

 

 

Any idea what this means? This was one of the disks that was recently rebuilt. Did the rebuild not work correctly?

 

I'm currently running new permissions utility to see if that does anything.

See check disk file systems in my sig.

I got the "rebuild tree" message. Should I pursue that step next? I'm reading the part where it says this should only be completed by expert users.

 

These drives are solely used for media files. I have all the originals and can recreate the data if necessary, so more important to me would be a) getting a list of files off this disk to replace, and b) preserving the integrity of the remaining disks/file structure, if that makes any sense.

 

Any insight as to why this is happening so I can correct it in the future. To summarize, Disk 1 reported errors while Disk 17 was being rebuilt. Once 17 finished rebuilding, I replaced Disk 1 and rebuilt that. After 1 was rebuilt, I was unable to view the files on the entire rig from a computer terminal. Now Disk 1 is giving the rebuild tree message.

 

Thanks for all the help.

Have you run parity checks after each rebuild? If not, do one now. Then run with rebuild-tree.

I try to run one each time, but they usually get interrupted due to the false errors from the disks in the external 4 bay enclosure. For example, I shut the server down last night after running the check disk. At the time, there were no errors. I just started it up to run the parity check, and it showed drive16 and 17 were unformatted and errors on 15, 16, and 17. I restarted and now it's making me rebuild 15. As soon as that began, it showed errors in 15 & 16 again (log attached). So before I can even run rebuild-tree, I have to rebuild 15 (2-3 days), run a parity check (1-2 days), and hope there are no additional errors. I just can't seem to make any progress with this. For reference, drives 1-14 and parity are all in the main rig, while 15-18 are in the 4 drive bay (18 is not assigned).

unraid.txt

Got about 5 seconds into the rebuild and 15, 16, and 17 went "bad" again.

Unraid's rebuild process requires all remaining drives to be read perfectly in order to write the correct data to the missing drive. If you had any read errors during a rebuild, the rebuilt drive would have bad data. Every time you rebuild a drive with the current unreliable setup, you are probably losing data.

So what would be recommended in my situation? None of the disks have been rebuilt while there were errors reported. I don't believe any of the disks were actually bad, except for disk 1.  All the other times, the errors have been off of disks in the external case. I've done check disks on those drives and they were all fine. Plus if you look at the above screenshot, it's saying it did 17 billion reads on three drives in less than 5 seconds. Seems like it's some sort of hardware/cable issue, which I just can't figure out.

You said you have just added an external drive bay. Are all the disks reporting errors in this external bay, could it be your power supply is overloading?

I added a four drive bay about 6 months ago, and my errors started shortly thereafter. It could be a power supply issue, but I really would have no way of knowing. Is there a simple check for that?

No not really but you would need at least 400 watts for the type and amount of drives you have. Removing the 4 drive bay may be your answer but thats not going to be easy to do. Im now moving all my data to WD 6TB Reds and selling my old 3TB and 4TB ones, the more drives you have the more power draw, more SATA ports and more likelihood of something going wrong with so many drives protected by just one Parity.

If you didn't get any read errors before adding the 4 bay unit maybe thats whats causing your issues.

I completely agree that it's the 4 drive bay. I actually had another thread on that one and tried several suggestions to get it working, but just keep running into the same issues with the bay having read errors before I can complete the rebuild/parity checks. I always had a gut feeling it was some data error (either cables or something else) and not a defective drive (3 drives would all show with the exact same number of read/write/errors for example). This was verified when I ran disk scans which all showed the drives were healthy.

 

I actually think I stumbled upon something last night while trying to go through all this. It might just be anecdotal, but it seemed that every time I unplugged the cable from the 4 drive bay and plugged it back in before starting up the server, it would give the server better odds of not showing errors. After I got errors at one point, I restarted the server and it immediately had errors again and 100+ billion read/writes. I did that cycle a few times with the same results before unplugging the cable, after which it came back up and worked fine. I'm currently 80% through the most recent data rebuild. I've replaced the cable previously, so that's not it, but could it be the cards on either the server or the drive bay?

 

If I can ever get this back running reliably again, I have plans to upgrade to 6TB drives in the server to replace the drive bay completely. I think I would just need 4 drives to completely replace the current 4tb parity drive and 3 3tb data drives and could migrate all data over. First, I have to get it running reliably though.

I'm currently 80% through the most recent data rebuild.

So much for that. Just had some errors pop up with just a few hours left. 512 on disk 15, 256 on disk 16, 256 on disk 17. I stopped the array and then 15 was showing "not installed" and 16 and 17 were "missing". Restarted the array and 16 and 17 are there again and 15 has to be rebuilt.

We all do things differently but my data is very important and I have off site backups. I would remove the 4 bay and offending drives, build a new parity for the rest and start to migrate to larger drives and then copy the data or whats left of it off the 4 bay unit. The 4 bay seems to be adding lots of problems but then again if the discs are old they could be the reason. You may lose data and you may not but you must have reliable hardware and the failure stats with less drives but larger capacity are far lower than many data drives and one parity. E-bay was my friend, I had no problems getting rid of my 6 x 3TB and 2 x 4TB drives, just two more 6TB's to preclear, a few more 3TB drives to sell and Im finished.

We all do things differently but my data is very important and I have off site backups. I would remove the 4 bay and offending drives, build a new parity for the rest and start to migrate to larger drives and then copy the data or whats left of it off the 4 bay unit. The 4 bay seems to be adding lots of problems but then again if the discs are old they could be the reason. You may lose data and you may not but you must have reliable hardware and the failure stats with less drives but larger capacity are far lower than many data drives and one parity. E-bay was my friend, I had no problems getting rid of my 6 x 3TB and 2 x 4TB drives, just two more 6TB's to preclear, a few more 3TB drives to sell and Im finished.

The drives in the 4 bay are all new 4tb drives.

 

I didn't realize this was possible. I thought you had to have a valid parity before I could remove a drive, move data, upgrade drives, etc? Are you saying I could upgrade the parity to 6tb, remove the 4 bay enclosure, replace a couple 3tb drives with 6tb drives, and then move data from the 4 bay back over? I thought there was a special process that had to completed to downsize the server.

The parity is there to backup any of the individual data drives. The data on the drives won't change if shutdown the array and you pull a drive. If you have one drive fail the parity will restore that drive. If you have two data drives fail at the same time then you lose the data on those drives only. The data on all other drives is still there and won't change unless one of those drives fails or has errors.

In order to rebuild a failed drive from the parity it uses the information from all the data drives and the parity to calculate what was on the failed data drive, if its the parity that fails you just rebuild the parity again to a new parity drive. Only the failed drive or drives lose their data, all other data drives remain intact. Read the wiki. Unlike say RAID 5 unRAID doesn't stripe its information across all drives meaning if more than one data drive fails you lose all the data on all the drives.

Some users don't even have a parity drive just data drives showing as one large drive. This defeats the purpose of unRAID which gives you protection using just one parity drive to protect all the data drives in case one of them fails. But the more data drives you have, then the odds of two failing at once increases.

Thanks megalodon for the description of how unRAID functions. I'm fairly familiar with most of the "how" unRAID functions but this definitely isn't my area of expertise, so that helps with some of the "why" it does what it does, if that makes any sense. I'm here on these boards to attempt to get help from people much smarter than me in this area. I'm not a programmer, so instructions are very helpful.

 

So here's the latest. (it might be time for a new thread, since the original request has definitely morphed since the thread was started).

 

TL;DR recap from earlier in thread: 15 drive machine functioned fine until I added a 4 drive bay. Ever since then, the disks in the 4 drive bay have been giving numerous errors (false reads, as verified by smart reports). Plan is to upgrade enough drives in the main machine to move all the data from 4 drive bay over and completely remove that from the system, back to the original 15 drives.

 

Configuration is: Main unRAID computer: 15 drives, 1 4tb Parity, 14 3TB data.

                          4 drive bay: 3 drives, all 3 4TB data drives. (will use the empty bay to preclear new 6tb drives)

 

I finally got the system to rebuild a failed drive and complete a parity check without any errors. Unplugging the cable from the 4 bay unit after every shutdown and plugging it back in has definitely helped with this. No clue as to why.

 

I ran another check disk on disk 1 and it came up with 17 errors, same as before. I ran the rebuild tree last night and it looked like it was finishing up this morning. I take it I will be able to pull the logs and see which files are corrupt and need to be replaced? Since all my media files are iso, this should be fairly easy to replace these files.

 

Now, assuming the share folder still does not show under the windows directory (original reason for this thread), I should run the new permissions utility again? Hopefully this will be successful since disk 1 has had the rebuild tree completed.

 

I went out and picked up 2 new 6tb drives. I'm planning to preclear those, which I'm guessing will take a week or more since the 4tb drives typically took 4-5 days. Once that's done, I was going to replace the parity drive with one and let unRAID rebuild that. Now, what happens if one of the disks in the 4 drive bay issues errors during the rebuild? Do I just keep trying until successful? Is there any way to rebuild parity without having the 4 bay connected?

 

Once the parity is upgraded, I was going to replace one of the main 3tb drives with the other 6tb drive. After that's completed, I was planning on moving data off of the drives in 4 drive bay. Each of those 3 disks has around 1.5-2 tb of data on it, so I would be able to move enough to completely free up one 4tb drive, possibly 2. Is there a recommended method of moving data? Is it as simple as opening up those drives in windows explorer and moving from one folder to another?

 

After the second 6tb drive is in service, and the data has been moved from one of the 4tb drives, I was going to move the now empty 4tb drive to the main rig to replace an existing 3tb drive. Same as above, I would move the remaining data from one of the two 4tb drives left to the now upgraded 4tb drive in the main machine. Repeat this until all data has been moved over.

 

Of course, there will be parity checks between every step, and I need to read a couple other threads on downsizing the number of drives, but is there anything else I should be aware of with this plan? Does it sound like the best route to fix these problems?

Running the rebuild tree helped make the share folder visible again. But sure enough, the parity sync failed when one of the drives in the 4 drive bay failed. Will start new thread and mark this solved.