Trying not to panic but am very concerned. I went to watch some content tonight and Plex is showing nothing is available. I hopped on the Unraid console and browsed my share and all of the folders are showing empty. I must of 5 or 6 TB of content that is missing. The last scheduled parity check ran last night with millions of errors. When I look at the parity logs it doesn't show last night's scan.

 

Looking for guidance,, thanks in advance.

Please post your diagnostics: Tools -> Diagnostics

Here you go.

 

A couple of updates.  i thought a restart of the server might help. After bringing down the array it was showing that two drives had failed - red circle. Not sure what the odds of that are, fairly unlikely. i reset anyway and everything came back on line. I also noticed that my older cache drive is running extremely hot, sometimes reaching 50 degrees. I wonder if the heat inside the case is too much and the system is acting up? I'm also noticing our power is a lot less reliable these days. I don't have the server on a UPS so also wondering if the abrupt power down could be causing problems as well.

 

Look forward to your diagnosis

tower-diagnostics-20180717-0526.zip

3 minutes ago, Gouie said:

I also noticed that my older cache drive is running extremely hot, sometimes reaching 50 degrees. I wonder if the heat inside the case is too much and the system is acting up?

 

I would start with opening up the case and blowing all of the dust and dirt out.  (Be sure to do it in a non-living space as it can make a real mess place!)  Then check to see that all of the fans are running.  Make sure that the fan system is setup to force air flow over the drives first and then help pull it out of the case after that. (If no fans in the front of the case, then all fans push air out of the case. 

If it happens again grab the diagnostics before rebooting.

A UPS would be a wise investment, abrupt power failures are never good and can lead to data loss or corruption. Do you have a backup for your data on your unRAID server?

Sorry everyone, appreciate the prompt response then life got really hectic. I've blown the case out and confirmed that all fans are spinning fine. I think I need to take a good look at air flow, push vs. pull to see that the drives are getting proper cooling. I ran into the problem again today. Took the array offline because things were behaving strangely; I couldn't run an LS on the user folder from the console, came back with an error. Once I took the array offline it showed that two drives were bad. I've reset the server a couple of times now and it still shows the 2 drives are bad. Not good. A UPS is in my future and to answer your question no, I don't have a backup of the data - I figured that a redundant RAID system would be enough.

6 minutes ago, Gouie said:

I figured that a redundant RAID system would be enough.

RAID or unRAID is never a substitute for good backups. Everything doesn't have to be backed up, but important and irreplaceable data must have backups. Plenty of ways to lose data besides bad disks.

 

Post new diagnostics.

1 hour ago, Gouie said:

I figured that a redundant RAID system would be enough.

 

RAID is about availability - that a machine can continue to give access to files after a disk has failed.

 

R stands for 'Redundant' - so more than one way to produce the stored data.

 

But a RAID will never replace the needs for a backup. On one hand, you may lose more disks than the number of parity disks - and a broken PSU can kill every disk in the system at the same time. On the other hand, disk redundancy can't help with file deletes or file overwrites or broken file systems.

Gotcha, I just figured the odds of tossing more than one drive before a guy had a chance to get a replacement were pretty low. 

 

Couple of questions now:

 

Backup; what are folks using? 20+ TB makes backup a little complicated. 

 

Failed drive; shouldn’t the console give indication of a failed drive when it happens? I’ve had to take the array offline for it to highlight a failed drive. 

 

Update; I downed the server again and gave it a real thorough clean. I’ve also left the case off for now. I have 8 drives, 6 attached to onboard ports with the other 2 attached to a SATA expansion card. Got me thinking perhaps the card is wonky. I lifted and reset it as well as changing the drives connected to it. If 2 drives fail again and they aren’t the same ones it will indicate a bad card. 

 

Thanks a a million for the help everyone!

 

49 minutes ago, Gouie said:

I’ve also left the case off for now.

 

Leaving the case of will generally cause things (including the drives) to run hotter!    Most cases are designed so that the air enters through the front, passing over the drives to cool them, and then through the Motherboard/addin-card area then out the back/rear top of the server.  Basically, you are depending a forced air flow through the case to keep temperatures under control.  

Argh! I’ll put the case back on. This case is a little different than most, designed as a home theatre case. It looks like an AV receiver. Lots of fans under, along the side and the back. Generally speaking the drives stay within range even when running a big copy - usually under 40C. The cache drive is the real problem, North of 45C at times. Think I’ll replace it with a flash. 

 

If I were a gambling man I’d bet in the add-on card as the culprit. I’ll be sure to report back with my findings. 

1 hour ago, Gouie said:

If I were a gambling man I’d bet in the add-on card as the culprit. I’ll be sure to report back with my findings. 

 

I had one of the two port versions (I think they all use  the same SI controller chip) after it had been in use for more than two years.  I found this quite surprising as a failure of this type is not typical for a middle aged part.  That is the reason that I now have two of the LSI based controllers in my two systems...  

Wanted to provide a quick update to close the issue. The system has been stable since re-seating the SATA card - looking like that was the issue. Regarding the heat, I tore into the case one more time and believe I've found that problem as well. The case came with 5 fans installed from the manufacture, 3 under the drive bay and 2 at the rear. Lots of air flow, the only problem was they installed all 5 to be drawing into the case, no venting. I changed the 2 at the rear to exhaust and so far so good.