My Working Configuration (Asus M3N-HD)


Recommended Posts

Hi,

 

I've just built an unRAID server so thought I would share the hardware I'm using as setup wise it has worked flawlessly.  Obviously I need to leave it for a couple of weeks to make sure it remains stable but the fact it is booting fine is a good sign :)

 

Motherboard: ASUS M3N-HD/HDMI nForce 750a SLI - 2x 16xPCI-E, 2x 1xPCI-E, 6-SATA with USB Booting, HDMI and VGA output (Onboard Graphics)

CPU: AMD 4850e X2

RAM: Corsair 2x1GB 4-4-4-12

Hard Drives: 2x 1TB WD Caviar Green, 1x 500GB WD

USB Stick: Freecom DataBar 8GB

 

The case used was a Silverstone KL01 as a) looks nice b) provides accessible drive bays down the full front of the case (lots of room for Icy Dock caddys).  The Freecom USB stick isn't the fastest but thats not too important I guess.

 

I'm currently running unRAID 4.4.2 with BubbaRAID.  The only initial problem I had was that I couldnt get the Realtek NIC to work but that appears to be because I forgot to name the flash drive :)

Link to comment

Glad you have had success with this motherboard and configuration.

 

After the initial parity build (which unRAID calls a parity check but it really isn't), you need to run a manual parity check.  Some users have no problems with the build, but the check fails!  If that happens, it means that your have a problem, and the motherboard might not be working after all.

 

It's sort of like getting a new DVD burner, and burning a disk.  Would you burn a backup disk of your most precious digital photos and put it away for safe keeping assuming the burn was good?  No, you'd put it back in the computer and make sure that the files were there.  You might even compare each of the files with the source material to make sure that it worked perfectly.  Point is, you wouldn't trust it just because it said it burned.  Similarly, you can't trust unRAID's parity build just because it built.  You need to check it.

 

After the check finishes successfully, I'd recommend posting a syslog.  If there are any nasties one of the experts here may be able to point it out to you before you have a problem.  You should also take baseline smartctl reports (see the troubleshooting link in my sig for instructions to save a syslog and get smart reports).

Link to comment

Glad you have had success with this motherboard and configuration.

 

After the initial parity build (which unRAID calls a parity check but it really isn't), you need to run a manual parity check.  Some users have no problems with the build, but the check fails!  If that happens, it means that your have a problem, and the motherboard might not be working after all.

 

It's sort of like getting a new DVD burner, and burning a disk.  Would you burn a backup disk of your most precious digital photos and put it away for safe keeping assuming the burn was good?  No, you'd put it back in the computer and make sure that the files were there.  You might even compare each of the files with the source material to make sure that it worked perfectly.  Point is, you wouldn't trust it just because it said it burned.  Similarly, you can't trust unRAID's parity build just because it built.  You need to check it.

 

After the check finishes successfully, I'd recommend posting a syslog.  If there are any nasties one of the experts here may be able to point it out to you before you have a problem.  You should also take baseline smartctl reports (see the troubleshooting link in my sig for instructions to save a syslog and get smart reports).

 

I have only copied data from other sources at the moment onto it so i've got two copies still.  I'm going to run a full check overnight tonight and will then take a look at the logs tomorrow - cheers for the headsup on that.

Link to comment

It completed a parity check last night with 0 errors.

 

My syslog can be found at:

 

http://www.delimits.net/richard/syslog-2009-02-07.txt

 

The only thing I see in there that is worrying me - although it has recently entered the log is:

 

Feb 7 00:45:40 nas kernel: EXT3-fs warning: mounting fs with errors, running e2fsck is recommended

 

The smartctl reports for the three drives can be found in:

 

http://www.delimits.net/richard/smart-sda.txt

http://www.delimits.net/richard/smart-sdb.txt

http://www.delimits.net/richard/smart-sdc.txt

 

Some further information is that the system is doing a sync at about 55MB/sec and will also wake on lan using the magic packet. To make that happen I had to use 'ethtool -s eth0 wol g' to set the 'Wake-on' to 'g' however, upon rebooting it has gone back to 'b', should I place the ethtool command in my go script?

Link to comment

Your array is the picture of health!

 

You should feel confident that if a drive were to fail that you could recover.

 

I didn't realize you were running bubbaRAID.  In general I'd recommend people just use "stock" unRAID until they get their array setup and stable.  This just eliminates it as a possible issue IF a person is having a problem.  But you seem to be fine and no reason to go back.

 

You SMART reports each look perfect.  Nothing even remotely concerning there.  (Keep these reports as a baseline, if you have problems later it will be helpful to compare).

 

Your motherboard is Level1 verified.

Link to comment

Haha, a bit too eager perhaps in wanting to test out all of the little addons and what not, i've also got unMenu and the alternate Powerdown script active - both fully working.

 

I'm very happy with the unRAID system as a whole.  Once I've got some more money I will be looking to upgrade to running with a proper license and using a cache drive to improve performance.  I've got another 500, 2x 200 and a 250 to add into the array.  The only thing I would like to have is a more polished user interface - I don't know how easy this would be to do yet, I need to look into how the system works a bit more and how easy it is to change things.

 

The other thing is that nearly all of my 2GB of memory is used for caching, would increasing this improve anything?

 

 

Link to comment

Thanks for the tip RobJ. I have set the NCQ for all drives to 1 and then used the 'hdparm -tT' option as part of unMenu which gave the following:

 

/dev/sda: (1TB 32MB cache) - Parity drive

Timing cached reads:  2188 MB in  2.00 seconds = 1094.19 MB/sec

Timing buffered disk reads:  222 MB in  3.14 seconds =  70.69 MB/sec

 

/dev/sdb: (500GB drive)

Timing cached reads:  2134 MB in  2.00 seconds = 1067.61 MB/sec

Timing buffered disk reads:  152 MB in  4.12 seconds =  36.91 MB/sec

 

/dev/sdc: (1TB drive)

Timing cached reads:  2180 MB in  2.00 seconds = 1090.01 MB/sec

Timing buffered disk reads:  242 MB in  3.01 seconds =  80.38 MB/sec

 

/dev/sdd: (USB pen drive afaik)

Timing cached reads:  2060 MB in  2.00 seconds = 1030.47 MB/sec

Timing buffered disk reads:  54 MB in  3.02 seconds =  17.86 MB/sec

 

I will run a parity check overnight tonight and report back the average. Are there any other good ways to run speed tests? It seems that the read performance on my 500GB drive is half the 1TB drives.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.