unRAID stops responding, permissions issues

[Croaker]

Alright, so I'm running unRAID 5.05 on a system with a WD Red 3TB parity drive, 3 more WD Red 3TB data drives, a Seagate 7200.14 2TB drive and a 1TB Seagate 7200.12 as a cache drive. Full system specs at the bottom of the post.

 

After a period of time, sometimes a few minutes, sometimes a couple of days, the unRAID web interface will become non-responsive and time-out, and I will be unable to access sickbeard, CP, or Transmission. I can *usually* still TelNet into the server at this point for a powerdown, but not always. Sometimes Transmission will spit out issues claiming that it's trying to write files to a read-only location despite the fact that the cache drive is CHMOD'd as 777, tried running as both nobody and root with the same result. SickBeard will also have trouble post-processing downloaded items and fail to relocate them to their permanent location on the array off the cache drive.

 

I haven't installed anything new or modified my config in any way in a couple of months - just have Transmission, CouchPotato, and sickbeard doing their thing and downloading stuff which I'll end up watching through Plex. The problem seems to be getting worse. I've run a reiserfsck on all of the array drives and found no issues. I did the same check on the cache drive while the array was started in maintenance mode and it finished almost immediately with no issues, so I'm not sure I checked that one correctly (reiserfsck --check /dev/sde1 ).

 

Is it possible these problems are being caused by a failing cache drive? The drive is pretty old and does return some SMART test errors (End-to-End Error reads as Failed Now), but it's been returning that since I installed it and I haven't had these problems until recently. Any other recommendations on things I can check to try and track this issue down? I'm not particularly Linux savvy, but I'm tech-savvy in general and I can follow instructions.

 

 

System Overview

unRAID Version: unRAID Server Plus, Version 5.0.5

Motherboard: ASRock - FM2A75M-DGS R2.0

Processor: AMD A4-5300 APU with Radeon HD Graphics - 3.4 GHz

Cache: L1 CACHE = 96 kB (max. 96 kB)

L2 CACHE = 1024 kB (max. 1024 kB)

Memory: 2048 Module (max. 8 GB)

2048 MB = A1_BANK0, 667 MHz

Unknown = Unknown,

Network: eth0: 1000Mb/s - Full Duplex

 

 

Syslog: http://pastebin.com/aW4gtvVt

[dgaschk]

Any time a SMART report says FAILED NOW the disk is done. A disk may fail before SMART says FAILED but it never lasts longer.

[Croaker]

Noted. It's returned that result and run fine since I put it in several months ago, but I'll pick up a replacement cache drive in the near future. That being said, is it possible that this is the cause of all the problems I've been having lately or does it sound like I may have multiple issues here?

[DaleWilliams]

A couple of things in the syslog:

 

How did you delete the shares for 'TM' and 'Dropbox'? via the SF interface?

 

1. 'SimpleFeatures' is no longer supported, and, with v.5.0.x the SF web interface is wrong. Functionality is broken.

You might like 'Dynamix' which has become a popular replacement for SF. I think you'll find similar plugins to those you have.  I don't recall anyone posting that SF freezes the webgui, but it would seem to be a possibility. And the mess-up with user shares is consistent with my own SF+v5.0.x experience.

http://lime-technology.com/forum/index.php?topic=28927.msg279478#msg279478

 

2. You've a lot of plugins, and I didn't have time to go through them all, but you might check them for compatibility with 5.0. 

 

 

[Croaker]

Thanks for the info. I ended up getting reiserfsck done on the cache drive and it did indeed return as all kinds of messed up. I replaced the drive today. Also uninstalled SF and installed Dynamix. 

 

During preclear of the new cache drive, I encountered an error that killed connectivity to the server. Link is pasted below. I wonder if this has been what's happening when I lose connectivity with the server on other occasions and I just happened to catch it now?

 

http://pastebin.com/A4UEjje5

[DaleWilliams]

1. Don't throw out that cache drive you replaced...it may be good, and, if nothing else, can become a future spare. Run a SMART test on it, and if clear, just do a preclear on it.

 

2. There's a special thread for preclear questions...

Post your last question there for a faster response:

http://lime-technology.com/forum/index.php?topic=4068.0