System was built many months ago and was working fine until about a month ago.

It began with a movie simply stopping playback one night.  Unraid server stopped responding and needed rebooting.  parity check was no errors.

Since then problems occur after a few days.  Strange issues, Movies copied over to not show up but when I try copy again it asks to overwright.

A few days ago many parity errors.  Cleaned up and though OK, however tons more parity errors again.

I noticed parity drive shows no free space.  No write errors at all.  Temps claimed high on the parity drive (51 degrees) during a parity check but it does not feel hot at all, just very warm.

It was built with all parts in the recomended list, Gigabyte MB, Cooler master Centurion case with 2 ICY Dock 5in3 backplanes, celeron CPU and 2GB of RAM.  System is ALWAYS very cool to the touch.

 

Here is log if anyone could help that would be great!!

 

Thanks!

 

It is normal for the parity drive to show no space.  51C is high but probably not the cause of the issue.

 

Smarter people than I will soon chime in with more useful advice.

 

 

Bill

As Billped said, I'm surprised smarter people haven't chimed in yet. 

 

There are other posts on here that describe how to do smart checks on your drives.  That is the first thing I would do.  My gut reaction says that if you have smart errors on either your parity drive, or the drive the movie was on, it's probably time to replace that drive.  Of course keep in mind at all times that any other general computer hardware problem could be the culprit, backplane, sata cables, memory, sata controllers, could all be wigging out, but all things being equal it's probably a drive problem. 

Well until these smarter people show up, I'll offer a few comments:  we need to gather more information and we need more testing.  I'm rather sure that I have seen similar situations, and I'm not confident they were always completely resolved, but let's see what we can do.

 

The advice given was good, get SMART reports for every drive, and examine them for current issues, so we can eliminate all of the hard drives as a cause.  There are no apparent drive errors in the syslog at all.

 

This syslog begins with a rather large number of transactions replayed for every single drive, which is very unusual, and it is remotely possible that that is throwing off the parity check, since the playback and parity check are simultaneous.  The number of transactions to be replayed is much more than normally occurs after an improper shutdown.  So we need you to shutdown properly and reboot, make sure the system comes up correctly, and is not running a parity check, then initiate a parity check, and after awhile capture another syslog and attach it.  It would also be helpful if I could see any syslogs you may have that recorded the earlier series of parity errors.  I'd like to compare the series of parity errors for randomness or patterns.

 

Also, to rule out a remote possibility, run a memory test overnight.

Thanks for the help folks.

 

I will try those suggestions tonight. 

 

Until then, here are some updates:

I cannot complete a new parity check.  The system becomes unresponsive after a period of running the parity check and I eventually have to force a power off.  I can ping the server but I cannot open the web interface nor access any shares.  It does not boot the first time after that.  A 2nd reboot usually starts it up and then I go and cancel the parity check so that it does not freeze again.

Could it be that the parity drive is overheating in the 51 degree range and causing the issues during a parity check?  I can't see how I could add cooling to the drives since they are in a 5in3 backplane with zero space for anything.

51c is too hot. I know my 10,000 rpm drives shutdown at 55c. (They never reached that temp).

I would check all cables and reseat all drives, check all fans.

 

There's a part of me that wants you to do a reiserfsck on each of you drives after you unmount them.

To have files not be visible, yet having you validate overwrite seems to lead me to some form of corruption.

 

Let's see what the smarter people say 

Well, I feel dumb.  I was going to tell you to run reiserfsck tests, and completely forgot.  Run them on all of the data drives, but NOT the parity drive.  See the Check Disk Filesystems HowTo page.

 

He's right, that's too hot.  Make sure the fans are all working.  Five drives in a space for 3 is too close, without active airflow above, below, and between each of the drives.

That is crazy.  I copied kung fu panda to my unraid server the other day.  It did the same thing.  I had just assumed it was a glitch in the network cause it failed to copy. then when i tried it again it asked if i wanted to overwrite it even though it wasnt there. I said yes obviously cause apparently im an idiot because now the parity drive has a red light.  I stopped the array and restarted it thinking maybe, just maybe, it was just some sort of minor glitch.  Now model/serial no. i get giberish,  temp is *, size 0, Free - , reads 0, writes 288, and errors 288.  the hd is a 1.5 td seagate that is only a few months old.  I have it in an external drive with its own cooling fan.  It is connected to the server itself thru an esata cable and its been working fine since i built it a year ago.  Im up to ten tds and i was about to add another 1.5 tb seagate untill these shenanigans started.  What are all these tests you speak of?  Oh yea and buy the way when i stopped the array the parity drive turned blue so when i started it up again i picked rebuild array.

 

thanks in advance for any advice.

Just looked at my new egg order history.  I bought the drive 10-16-08.  Missed the 30 days by 1 freekin day

Ok well i switched out the hard drive and the external enclosure.  Same weirdness.  What do i do now???

I dont get why it would still say the parity is red.  Its a brand new drive and when i fired up the server again it never formated it. It just went right to writing and 288 errors.

 

I stopped the array and now i only have the start (unprotected) option or restore

 

What do i do?

 

This is killing me.  My kid is freekin out without being able to watch inspector gadget in the theater before bed.  I dont want to do anything until i know everything will be ok

Parity drives are never formatted.  They hold a collection of bits, not files.

 

If your parity drive is "red" you can try the "restore" button, or you can just "start" the array and let it calculate parity. (are you sure it is not "orange" indicating not yet calculated?)

 

If there are errors in writing to the new parity disk it will be taken off line and really be "red".  At that point, I'd suspect the cable, tray, or controller rather than the new (or old) drive.

 

Joe L.

The tray is new and i just tried a new cable and still red.  Under the parity drive section it says not installed and the rest is blank.  In devices everything checks out so i guess im gonna try a new port, if i even have one. 

 

One other question, I just replaced the cable(esata) and hit refresh.  Do i have to shut it down completely for it to get a good reading on whats going on or is that something that should just pop up?

 

 

thanks for the input buy the way.  I really appreciate it.  It seems like people have been on strike or something lately from answering forum posts.  Im on other forums and it seems like the last few weeks many questions have simply been unanswered.

Are you plugging the eSATA cable in without powering down?  (It sounds like you are.)  As far as I know, unRAID does not detect eSATA drives that are hot-plugged.  The only drives I know that work that way are USB drives, but they are not able to be part of the unRAID array at this time.

 

Regardless, "stop the array, go to the devices page, un-assign the parity drive, reboot, stop the array once more(if it starts), go to the devices page, assign the parity drive and then start the array once more.  It should build parity as long as the drive is writable.

 

Before you reboot, a copy of your current syslog would help us to diagnose what is happening.

 

Joe L.

how do i copy the syslog?

 

Im sorry.  I really have know idea what the hell im doing, i just float through and some how get by.

ok i did it.  Its attached and its huge.

 

thanks

Now when i reboot the server it isnt seeing my parity drive at all.  I cant even assign it to a disk, and i switched ports.

Now when i reboot the server it isnt seeing my parity drive at all.  I cant even assign it to a disk, and i switched ports.

I would first suspect the cable, or a loose connection...   Of course, since it is a new drive, it could have died an early death.

 

You never answered my earlier question... are you powering down when changing connections, plugging/un-plugging in the drive, etc?

 

Joe L.

Yes, ive tried everything at this point.  Ive tried switching the cables, switching the sata ports, i have tried three different external enclosures (luckily i have about seven of them), and ive tried doing it with the server on and off.  I have also tried all of the above with both drives.  Both seagate 1.5 tb.  One is a month old and 1 is fresh out of the package.

 

this sucks!

Ok well i have no clue why but i just tried putting the original drive back (in a new enclosure and new wire) in the original port and it worked!!!!!

its doing the parity sync now.  Honestly i dont get why.  I made sure everything was plugged in correctly, nothing was lose, I switched everything around and tried different parts, i tried it a hundred different ways from Tuesday and now for some reason it works.  I dont get it but thanks for all the help.  Hopefully i wont get any errors or anything and i can add that other disk cause im out of space.

 

Thanks again

Just thought I would chime in with an update.

It took me a while as I have been very busy in my professional life.

 

During a reboot, I noticed the memory test option and decided to run it.  No idea why, it was just an instinct kind of thing.

Turned out one of the two 1GB modules was defective.  Pulled it and things have been working great for two weeks straight.

 

Thanks for all your help!

Added a link to here in the FAQ, here.  Feel free to edit or expand.