MSI K8N Diamond - Am i going to be flogging a dead horse?

[Kode]

I've been toying with building an unraid server for a long time, i've finally decided to start the process (it's coming up to my birthday so i can ask for parts off family).  I wanted to reuse some of the equipment i had lying around as much as possible.

 

So with that in mind i bought a new case on Thursday, it arrived on Friday, and i moved all of the bits out of my old gaming rig.  The gaming rig originally had an nForce4 MSI K8N Diamond motherboard, with an AMD FX-55 CPU and dual 7950GTX's in SLi.  I've removed the GTX's (got a PCI gfx card i can use for building it), this means that i have 6 SATA ports on the motherboard, 2 x PCI-e x16 slots and 1 x PCI-e x4 slot for future expansion, which i thought would be perfect, but have noticed that nforce4 motherboards arent recommended.

 

What are peoples suggestions here?  Is it a case of using the on board SATA ports that cause the problem, so if i start out using a Supermicro AOC-SASLP-MV8 which i believe is supported in the new build, that will solve my issues, this will still allow me to expand  to 24 drives, more than possible in a protected array currently.  Or is the whole board in general just a write off?

[terrastrife]

nvsata has come a long way since the wiki was written.

 

i run an nforce 500 which is really just a rebadged nforce4 ultra (some nforce500 boards still carry the nforce4 range names) with no problems with the onboard sata.

 

grab the free version and give it a go first.

[Kode]

Thanks for replying terrastrife, i already have the free version installed, i booted it up last night and it got to the tower login stage, didnt have time to do anything else.  What in particular do i need to look out for to make sure its suitable? 

[Kaygee]

Corruption issues with the onboard sata controller.

[Kode]

I'm sorry, i badly worded that, i understand that curruption issues with the onboard sata controller have been flagged from reading the wiki, my question really was where do i need to go to check if i'm having these issues, are there any commands i need to run, any logs i need to study to look for problems?

[ohlwiler]

I think the easiest way is to install TeraCopy on your PC, check the preference "Always test after copy" and start copying some big files to the array  (5-20 G.). TeraCopy will do a hash of the original file and the copy on  the unraid server. It is also a very nice way to copy large amounts of data. It is very popular with the unraid crowd.

[Kode]

Ok, that sounds like a plan, thank you, is the corruption issue a problem on the initial copy or over time though?  Because if its on the initial copy, doing a test straight after would show issues and i would know i had no choice but to get a new motherboard, however, if its a problem with corruption occuring later then its going to be a lot harder to verify i have a problem.

 

I currently have 4 x 1TB drives in a windows spanned volume so if a single drive fails potentially the whole lot is gone, i have 1TB free, so my plan was to buy 3 x 1.5TB drives, copy all my existing data over then add the 4 extra drives.  However, because im not in a big rush to add the extra drives after the copy, i can keep them as back up for a while (though i have put off adding more data to the volume so i dont have to get 4 x 1.5TB drives instead of 3, so id like to add them as soon as possible) but i just need to know what to look for.

 

Any help is greatly appreciated as is the help provided so far.

[Joe L.]

Ok, that sounds like a plan, thank you, is the corruption issue a problem on the initial copy or over time though?  Because if its on the initial copy, doing a test straight after would show issues and i would know i had no choice but to get a new motherboard, however, if its a problem with corruption occuring later then its going to be a lot harder to verify i have a problem.

 

I currently have 4 x 1TB drives in a windows spanned volume so if a single drive fails potentially the whole lot is gone, i have 1TB free, so my plan was to buy 3 x 1.5TB drives, copy all my existing data over then add the 4 extra drives.  However, because im not in a big rush to add the extra drives after the copy, i can keep them as back up for a while (though i have put off adding more data to the volume so i dont have to get 4 x 1.5TB drives instead of 3, so id like to add them as soon as possible) but i just need to know what to look for.

 

Any help is greatly appreciated as is the help provided so far.

The errors showed themselves when repeated parity checks resulted in random occasional parity errors.  So, write your files to the unRAID array, verify their checksums, then run parity check after parity check.  You should never* have a parity error to correct. 

 

Joe L.

* unless you power down unexpectedly and don't cleanly stop the array first, in which case a parity check is automatically performed upon power restoral.  It is normal to occasionally find a few "parity" errors near the start of the disk, especially if write activity had occurred recently and the disk buffer cache had not been completely flushed to the physical disks..  They represent housekeeping bits (time-stamps?) in the file-system.

 

 

 

[Kode]

Thanks Joe, so if i use TeraCopy to transfer the 3TB to the Unraid array, check it after that, then run say 10 parity checks, that should give me a fairly good indication of wether i'm going to have issues with this motherboard?

[Joe L.]

Thanks Joe, so if i use TeraCopy to transfer the 3TB to the Unraid array, check it after that, then run say 10 parity checks, that should give me a fairly good indication of wether i'm going to have issues with this motherboard?

yes

[Kode]

I've finally started the build, 3 x 1.5TB Samsung F3 HD153WI drives arrived today. Unfortunately the serial numbers end 796, 798 and 800, so less than ideal, but we'll see.

 

The motherboard has 4 SATA ports and Silicon Image provides a further 2 (so 6 in total) i have put data disks on SATA1 and SATA2 and the parity on SATA5 (the first Silicon Image port) dont know if this was a good idea or not, but seemed like one at the time.

 

I have installed unmenu, then i added openssh and screen.

 

I've created 3 screens each doing a preclear.

 

What are normal temperatures to be seeing during a preclear?

 

 

[Joe L.]

What are normal temperatures to be seeing during a preclear?

I usually see between 35 and 40C.

 

Joe L.

[queeg]

Temperatures depend on the airflow in your case.  But the preclear writes to the drive a lot so they can heat up.  In my case they run 5 to 7 degrees hotter than when idle.  But I have really great airflow.  Mine idle at or below 29c degrees.  Preclear around 35c to 37c degrees.

[Kode]

shit

[Kode]

Disk 1: Disk Temperature: 54C, Elapsed Time:  1:35:34

Disk 2: Disk Temperature: 49C, Elapsed Time:  1:34:17

Disk 3: Disk Temperature: 46C, Elapsed Time:  1:33:39

 

I have just put all the panels back on see if it directs airflow a bit better, i have the fan attached to the system fan header, cuz i dont have any 3 pin to molex connectors here atm.

 

Using this http://www.scan.co.uk/Products/SilverStone-CFP52B-Cooling-front-Panel-converts-3x525-into-4x-35-bays-Hot-Swap-black-RoHS

[aiden]

shit

 

Well said  

 

Disk 1: Disk Temperature: [glow=yellow,2,300]54C[/glow], Elapsed Time:  1:35:34

 

You know that's like 130 degrees F... that's way too hot for a drive.  Usually 55C is the top of the "operating" range for a hard drive, but you don't want to be there very long if you expect the drive to last.

 

As a comparison, with a 120mm fan right in front of the hard drive cage my drives average around 24 - 26 C at idle, and 30 - 32 C during a preclear.

[queeg]

That should be burning hot to your hand.  Is it?

[Kode]

bloody thing, the fan part wasnt pushed back far enough to start the fan, its now at a slight angle, but the fan is running at least, hopefully the temps will come down...

[Kode]

2 mins later and down to 48C and dropping, probably a close call there... lol, thanks guys

[queeg]

Put an led fan in and you'll alwasys know if it's running.  Plus it will look nice.

[Joe L.]

shit

I am assuming that is a technical term in your country.   

[Joe L.]

2 mins later and down to 48C and dropping, probably a close call there... lol, thanks guys

You can thank another forum member who suggested I add the drive temperature to the display as the disk was cleared. 

 

He too nearly roasted a disk.

 

Let's hope you get down below 40C.

 

Joe L.

[Kode]

Im such a retard, i highlighted the disk temperature and pressed Ctrl + c.. thats the danger of ssh-ing into a linux box from a windows laptop, i went into windows mode, Ctrl + c obviously quit the process so i had to start it again for SDA, anyway, the temps are looking better

 

Disk 1: Disk Temperature: 31C, Elapsed Time:  2:15:45 (Just before ctrl + c)

Disk 2: Disk Temperature: 30C, Elapsed Time:  2:16:38

Disk 3: Disk Temperature: 31C, Elapsed Time:  2:15:34

[Kode]

This is what im using

As you can see below there is a connector on both

When they contact the fan spins

There are tabs on either side to stop it being pushed too far, but i've had to go past these on the left for it to contact, you can see the tabs on the right hand side, but not the left now.

As you can see i cant put the rack any further forward its as far forward as it can go and be screwed in

[Kode]

When the fan is spinning it seems pretty good, temps are down to:

 

Disk 1: Disk Temperature: 26C, Elapsed Time:  0:23:10

Disk 2: Disk Temperature: 25C, Elapsed Time:  2:38:41

Disk 3: Disk Temperature: 26C, Elapsed Time:  2:37:59

 

And yeah, it wasn't a technical term, it was an acronym, for... so hot its ... a word beginning with t, cant think of a suitable one