Help with Smart Report Fail and fan speeds

[jakebake]

I haven't used my unraid server in months if not a year and I just reset it up with OS 6. I have always had one 2tb parity drive and one 2tb data drive. I just recently added a ssd cache drive and a WD Red 2tb data drive. I ran a smart report on my first data drive, however it only gets to about 10% and stops. I think it is going to fail but I do not know when or how bad it is. I attached the SMART report. If the drive is failing then what is my best course of action? I am not sure if I want to buy a new drive right now so I am guessing I can just move all the data to the new WD Red I added?

 

My other question is I only had a cpu fan and an exhaust fan in the back of the case. When resetting up the server I noticed that the exhaust fan is no longer working. I added a yate loon fan in the back and the front of the case because they were the only extra fans I had. The problem is that they are super loud. Is there a way to control fan speed in Unraid? My motherboard is a zotac nvidia MCP61.

 

thanks!

ST32000542AS_5XW28095-20160102-0021.txt

[JorgeB]

Disk has 3000+ reallocated sectors, SMART is failing,  replace it ASAP 

[JorgeB]

For fan speed control you can use Dynamix System Autofan

 

http://lime-technology.com/forum/index.php?topic=36543.msg339559#msg339559

 

[jakebake]

Ok thank you! Is there a reason this could have happened or it just happens? I've never had a hdd fail on me before. Is my best option to just copy the data from it onto the new drive and then replace it when I can or should I take it out and rebuild to the new drive from parity?

[JorgeB]

It just happens, if you trust your parity I would rebuilt it, anything goes wrong you can try copying data from the failed disk.

[jakebake]

Well crap it looks like my parity disk has the same problem. Is this because I just did a parity from the other drive recently? attached is the smart report for the parity drive

ST32000542AS_5XW23283-20160102-0118.txt

[JorgeB]

Yep, parity is also failing,  it has nothing to do with the other disk, but it's not normal for two disks to fail at the same time, if they were close to each other and without proper cooling they can overheat and cause the issues you're having.

[jakebake]

What is an OK temperature for the drives to be running at? And since I just noticed that the fan that was in the case was broken then the drives not getting enough cooling could be the cause. I am pretty sure that months ago when I stopped using the server the drives might have had problems.

 

So what is my best course of action here? I guess I do not understand rebuilding vs just moving everything from the first 2tb disk to the new one I just added. Or at least just moving everything that is super important to the drive I know works just  in case.

[JorgeB]

I like to keep my disks under 40C, but critical temp is usually >55C.

 

Note that although two disks failing together is unusual it’s also possible, especially with disks manufactured together, like I assume those Seagates were since they have similar serials, I’ve seen more than once disks like that fail together or within a few days of each other.

 

If the error counter for both parity and disk1 on main WebGUI page are at zero and you have a spare disk you could still try a rebuild, as bad sectores were in theory reallocated and replaced, but with so many I’d be surprised if it completes without errors.

 

If the rebuild fails or you prefer to skip it it’s probably better to do a new config with just both data disks and try to copy all data to the other disk, I would do this so that the bad parity disk won’t cause issues during the copy.

 

[trurl]

Do you even need that data? You haven't accessed it

...in months if not a year...

[jakebake]

I like to keep my disks under 40C, but critical temp is usually >55C.

 

Note that although two disks failing together is unusual it’s also possible, especially with disks manufactured together, like I assume those Seagates were since they have similar serials, I’ve seen more than once disks like that fail together or within a few days of each other.

 

If the error counter for both parity and disk1 on main WebGUI page are at zero and you have a spare disk you could still try a rebuild, as bad sectores were in theory reallocated and replaced, but with so many I’d be surprised if it completes without errors.

 

If the rebuild fails or you prefer to skip it it’s probably better to do a new config with just both data disks and try to copy all data to the other disk, I would do this so that the bad parity disk won’t cause issues during the copy.

 

I am confused by your last sentence? I ma not sure if I trust my parity so copying the date might be best. For now I am just going to buy another WD RED 2TB drive and use it as parity and use my other 2tb red as data. Also, buying 4tb is more cost effective than buying two 2tb drives. I know parity needs to be bigger or equal to all the other drives. If I wanted to buy the 4tb drive now then I would have a 2tb and a 4tb drive. Is there a way to use 2tb of the drive for parity and the rest for data? 

[jakebake]

Do you even need that data? You haven't accessed it

...in months if not a year...

 

The data is a good amount of movies, some TV shows, and some pictures and other data. I haven't  used it in awhile because wasn't watching much or just used netflix, etc. I would get over losing all the movies, but I would definitely rather keep them.

 

[JorgeB]

Is there a way to use 2tb of the drive for parity and the rest for data?

 

No that’s not possible, you could use a 4tb drive for parity, even if all data disks are smaller, so in the future you could upgrade current 2tb disks or add more 4tb disks.

 

In the last sentence I was suggesting that since your current parity drive can have bad sectors, it’s probably best to not use it during the copy, if you plan to buy a new parity drive, be it 2 or 4tb I’d suggest doing this:

 

-Create new array with new parity plus good wd red drive.

-Mount other data disk with unassigned devices plugins and try to copy all data to new array.

 

 

 

[jakebake]

Alright so I bought a new WD Red 2tb drive today. So for now I just have the two WD Red 2tb drives the replace my old parity and data drives. I took out my parity drive and put the new WD Red in its place. So with unraid powered up but not started I have Parity Missing and then the old disk 1 and disk 2 as the other WD Red 2tb drive I added before. Should I assign the new WD as parity and then start the array and transfer data from the old disk 1 to disk 2? I was confused by "Mount other data disk with unassigned devices plugins and try to copy all data to new array" are you saying to put the new WD red as parity and leave the other WD Red disk and then make disk one "No device" and boot up?

[JorgeB]

You can assign the new disk to parity and start array to rebuild it, but if disk1 has bad sectors there are going to be errors.

 

Another way of doing it without using the UD plugin, go to tools and do a new config, choose the old WD red as disk1, bad disk as disk2, no parity yet, start array and copy everything you can from disk2 to disk1.

Ignore this unnecessarily complicated way, was missing my morning coffee, to do this option just start the array without assigning parity and copy all data from disk1 to disk2, then do the new config.

 

When copy is over stop array and do another new config, this time with new and empty WD red as parity and old WD red as disk1, double check disks selection is correct and start array to begin parity sync.

 

[jakebake]

Sorry for asking so many questions but really confused at this point....

I started the array with no disk assigned to parity and the bad disk 1 in disk 1 still and new WD RED 2tb with nothing on it in disk 2. I then go to windows network drive and go to my unraid. I am trying to copy or cut all the files in folder "disk 1" to "disk 2" but it keeps sayingit is moving 12 TB of data and there is not enough room on disk 2. Idk how this is possible as disk 1 is only a 2tb drive

[JorgeB]

That sound like file system corruption, see Check Disk Filesystems and follow intructions for the filsystem you're using.

 

Since you’re going to remove disk1 you may want to try to copy data another way before fixing the file system, as depending on how bad it is it can lead to data loss.

 

E.g., try to copy with rsync, if it also fails check fs first:

 

rsync -avP /mnt/disk1/ /mnt/disk2/