Newbie Needs Help with First Unraid Build

Hello all, I have decided to jump right in and try to setup an Unraid server, and I was hoping to get some advice.

First, let me describe what my current setup is, and what my needs for this server will be. I currently have a Zotac HD-ID11 in the living room that runs XBMC, and has a wired connection to my Gigabit router (Linksys E2000). Currently, the Zotac reads media from a pair of external USB drives (1 x 2TB and 1 x 1TB). In the office, I have my main Desktop computer (Gigabyte GA-P35-DS3P, Core2duo E6750, Antec P182) with 3 HDDs (1 x 2TB, 1 x 1.5TB, 1 x 750GB) in addition to the main drive with the OS. This is what I use for downloads, etc. This computer is hooked into my home network via a wireless bridge, as a wired connection is not possible (or at least I'm not going to pursue trying to wire the house at this time).

Currently, in order to get new media in a place where I can watch it on my living room Zotac box, I usually unplug one of the USB drives from the Zotac, hook it up to my Desktop, and synchronize it manually. So painful! Synchronizing it over Wifi is just impractically slow, as I am usually moving around large HD movies. My current system does at least provide some redundancy, as everything is backed up on my Desktop. However, an unRaid system would obviously provide me with much more capacity.    

So, my plan is to build an unRaid server that will be tucked away in the living room, with a wired connection to my router to serve the Zotac box. Since it will be in the living room, a relatively quiet system is obviously preferred. I would also like to be able to use it for Usenet downloads (I currently use alt.binz on Windows, but it looks like sabnzbd could provide the same functionality?), as well as the occasional torrent, par2 repairing, unraring, and of course serving 1080p videos over the network. Here is my preliminary list of hardware to meet these needs:

Cooler Master Centurion 590 (seems like the de facto standard case)

Asus M4A78L-M

Corsair CMPSU-550VX

AMD Sempron 140 Sargas 2.7GHz

Kingston 2GB DDR2 SDRAM DDR2 1066

Icy Dock (Starting out with 2)

Adaptec 2241000-R PCI Express x4 4-port SATA adapter (Starting with one, knowing that I can expand with another one down the road)

And I will start out with 5 drives (2 x 2TB, 1 x 1.5TB, 1 x 1TB, 1 x 750GB), but I'd like to have the capability to expand to 10-12.

Is the Adaptec SATA PCI card a good choice? Or would it be better to go with something like the SUPERMICRO AOC-SAT2-MV8 or Supermicro Add-on Card AOC-SASLP-MV8 instead?

Would this be enough horsepower to meet my needs above? Are there any glaring issues that I should re-evaluate?

Thanks in advance for everybody's help!

What you linked to is a normal PCI card, and NOT a PCI Express card. They are completely different. In this day and age, you should avoid using PCI SATA cards. That is unless you do not mind having slower performance for parity checks, parity builds, failed drive conditions, failed drive rebuilds, and simultaneous use of those HDDs.

Oops, looks like I had the wrong hyperlink. It's fixed now. My intent was to show the PCI Express x4 card.

I do not know if that Adaptec card is supported.

The SuperMicro AOC-SASLP-MV8 card is supported and widely used by many users. It's a better value as it provides 8 SATA ports instead of 4. The other SuperMicro card is for server-based motherboards which offer PCI-X. Do not confuse PCI-X with PCI-Express.

I'll second the AOC-SASLP-MV8 - it is well vetted and a much better value than the one you linked (roughly the same price for 8 ports instead of 4).  Keep in mind that you will also need breakout cables for that card, see this thread for more info.

I would also recommend buying a single stick of 2 GB RAM instead of two 1 GB sticks.  While it will be slightly slower (single channel instead of dual channel), it allows for more expandability in the future.  2 GB should be sufficient for the uses you've outlined, but if you ever wanted your server to do more and upgrade to 4 GBs, you only have to buy another 2 GB stick of RAM and install it, you don't have to replace the whole set.  Also, the RAM speed doesn't matter much for unRAID, DDR2 800 or 667 is fine.  If you find faster RAM for about the same price then go for it, otherwise don't waste your money.

Other than those two things, I think the rest of your parts look good.  You have chosen a good motherboard and good quality parts all around.  You potentially may need a faster or dual core CPU for all the add-on stuff you've mentioned, but I'm not really sure, since I don't do any of that stuff with my server.

Also make sure you are using Cat5e or Cat6 network cable for your whole network.

Rajahal,

Thanks very much for the response. I've seen many of your posts on these forums, and have a lot of respect for your opinion. I'll definitely go with the AOC-SASLP-MV8, and I'll look into a 2GB stick of 667 or 800 RAM.

I read that post on the breakout cables, but I didn't quite follow it. Forgive the stupid question, but are these breakout cables just SAS to SATA cables? I'm not familiar with the SAS interface, but is this basically needed to connect the IcyDocks (which appear to have regular SATA plugs) to the SAS plugs on the AOC-SASLP-MV8? If so, looks like I'll be needing Forward Breakout Cables, correct?

I was looking at these. Is 20 bucks reasonable for one of these? Seems a little pricey to me, but if anyone has a better suggestion of where to get them I would be grateful.

Thanks again!

EDIT: I just noticed that each of those breakout cables can connect to 2 drives, so maybe it's more reasonable than I thought.

Yes, they are just SAS-SATA.  The confusing part is that they aren't two way cables, like regular SATA cables are.  You need to decide if you need forward or reverse cables for your situation.

Luckily you have it right, you'll need Forward breakout cables.  $20 is reasonable.  The cheapest I've seen them is $15 with free shipping.  You may be able to save a few bucks by ordering them from another vendor, for example eWiz (I just entered the cable's model number into google shopping to find that).  Also, the shortest length (0.5 meters) will be plenty long enough for your CM 590 case.

Actually, each breakout cable connects to 4 drives.  Since you will probably have to pay shipping for the cables, I would order the total amount that you could possibly need in one go (which is 2 in your case).  

Thanks, I'll go ahead and order two of the 0.5m cables then.

I think I'm going to spend a little extra money and get the Athlon II X2 250 3.0 GHz Dual Core processor instead. It's only about 30 bucks more, and hopefully that will be enough to cover my needs.

The only downside is that it is 65W compared to my original choice that was 45W. Hopefully the extra power doesn't translate to too much extra heat and therefore noise, but I guess we'll see.

For all the add-ons you want to run, that's probably a good call.

Ok, I have now received all of my hardware, and the system is up and running. Unfortunately I'm running into some issues.

I currently only have a 1 TB WD10EACS and a 2 TB WD20EARS (with jumper on 7 and 8 installed) with no parity. My plan was to try and copy over the ~ 2.5 TB of media I have first, and then install a parity drive. So I hooked up my external NTFS USB 2 TB Samsung drive, and followed the instructions here to mount and copy the files over to the EARS drive. The problem is that the transfer rate is only about 1 MB/sec! After about an hour of that, I stopped it and tried copying to the EACS drive instead, thinking that perhaps the EARS drive was the problem. However, I had the same results.

To top it all off, I'm getting some error messages during boot up. I don't remember the exact message, but it was similar to the errors discussed in this post. I had similar errors on the EARS drive, and I specifically remember seeing "failed command: READ DMA EXT" and ": EH complete" over and over for about 5 minutes before it would eventually boot up. Could this be related to the fact that it was previously formatted without the 7/8 jumper?

Below are pastebins of my syslog, as well as a smart report for each drive. If anybody has any insight here, I would really appreciate. I am starting to get really discouraged with all of this.

Syslog

SMART Report - WD10EACS

SMART Report - WD20EARS

To top it all off, I'm getting some error messages during boot up. I don't remember the exact message, but it was similar to the errors discussed in this post. I had similar errors on the EARS drive, and I specifically remember seeing "failed command: READ DMA EXT" and ": EH complete" over and over for about 5 minutes before it would eventually boot up. Could this be related to the fact that it was previously formatted without the 7/8 jumper?

Yes.

Lol ok. Is this a pretty common problem then? Is there anyway to fix this, e.g. running the preclear script?

Follow these steps.

That's only really necessary if you care about saving data on the drive.  If the drives are still empty, then just add the jumper and either preclear it again or let unRAID clear it.

Follow these steps.

 

That's only really necessary if you care about saving data on the drive.  If the drives are still empty, then just add the jumper and either preclear it again or let unRAID clear it.

I actually already had a jumper installed on pins 7 and 8. However, before installing the drive in my unraid server, I had previously formatted the drive (in NTFS) without the jumper installed. I'm hearing that the drive doesn't like that, but that seems odd to me ...

Any other suggestions?

By the way, I just started the preclear script on these two drives (they're still empty), and the EARS drive's read speed is hovering around 10 MB/s while the EACS is reading at about 70 MB/s.

Ok, after reading this post (as well as a few above it) it sounds like a lot of people are getting similar errors on EARS drives if they had previously used the unjumpered drive in Windows. However, it sounds like running the preclear script has fixed the issue for a lot of people, so hopefully that works for me (although it is painfully slow, after 14 hours I'm only at about 27 percent!

Back to my other problem: the slow transfer speeds when copying from an external USB drive. Does anybody have any thoughts as to why this is happening? Does the NTFS driver slow things down? Or could it have something to do with my BIOS settings? I followed the suggested BIOS settings to set USB emulation (I think that's what it was called) to "HDD" instead of Auto, or something along those lines. In any case, it is suspicious to me that I'm getting USB 1.1 speeds. I guess I'll make sure that USB 2.0 isn't disabled in the BIOS once my preclear is complete, but if anyone has any suggestions I would appreciate it. Otherwise I'll post back once preclear is complete.

after 14 hours I'm only at about 27 percent!

That is far to slow and it should not be taking that long.  A 2TB drive takes a while but it should be more than 30% done in 14 hours.

after 14 hours I'm only at about 27 percent!

 

That is far to slow and it should not be taking that long.  A 2TB drive takes a while but it should be more than 30% done in 14 hours.

Agreed. Like I said, my 1 TB drive is working at about 60-70 MB/s, and the 2 TB EARS is about 10 MB/s. I'm not sure why, but I posted syslogs and smart reports above if anybody has any suggestions.

Agreed. Like I said, my 1 TB drive is working at about 60-70 MB/s, and the 2 TB EARS is about 10 MB/s. I'm not sure why, but I posted syslogs and smart reports above if anybody has any suggestions.

I would say that you have been hit by the bug that a lot of other have.  If the drive did not have the jumper on it and was used it will have problems once the jumper is installed.

My suggestion is not even mess with the drive anymore.  Set up an RMA and get a new one.

Yeah I think I'm going to. I'm going to let the preclear finish first, as a last ditch effort (I'm going to be out of town for the weekend anyway, so hopefully it finishes before I get back), and if that doesn't fix it, I'll try to setup an RMA.

I've actually never returned a hard drive before. I bought it from Newegg, and it's been more than 30 days so I think I have to go through Western Digital directly. How does that work? Do they make you jump through hoops to prove that the drive is actually defective? Or can you pretty much just send it back with no questions asked?

Ok so my preclear has finished on one of my drives, the WD10EACS. I'm not sure what I'm looking for here, but can anyone tell me if this output would indicate a problem?

Disk Temperature: 32C, Elapsed Time:  19:40:40
===========================================================================
=                unRAID server Pre-Clear disk /dev/sda
=                       cycle 1 of 1
= Disk Pre-Clear-Read completed                                 DONE
= Step 1 of 10 - Copying zeros to first 2048k bytes             DONE
= Step 2 of 10 - Copying zeros to remainder of disk to clear it DONE
= Step 3 of 10 - Disk is now cleared from MBR onward.           DONE
= Step 4 of 10 - Clearing MBR bytes for partition 2,3 & 4       DONE
= Step 5 of 10 - Clearing MBR code area                         DONE
= Step 6 of 10 - Setting MBR signature bytes                    DONE
= Step 7 of 10 - Setting partition 1 to precleared state        DONE
= Step 8 of 10 - Notifying kernel we changed the partitioning   DONE
= Step 9 of 10 - Creating the /dev/disk/by* entries             DONE
= Step 10 of 10 - Testing if the clear has been successful.     DONE
= Disk Post-Clear-Read completed                                DONE
Disk Temperature: 32C, Elapsed Time:  19:42:09
============================================================================
==
== Disk /dev/sda has been successfully precleared
==
============================================================================
S.M.A.R.T. error count differences detected after pre-clear
note, some 'raw' values may change, but not be an indication of a problem
62,63c62,63
< 192 Power-Off_Retract_Count 0x0032   200   200   000    Old_age   Always       -       17
< 193 Load_Cycle_Count        0x0032   179   179   000    Old_age   Always       -       65996
---
> 192 Power-Off_Retract_Count 0x0032   200   200   000    Old_age   Always       -       18
> 193 Load_Cycle_Count        0x0032   179   179   000    Old_age   Always       -       65997
============================================================================
root@Tower:/boot#

As for my WD20EARS drive, it is chugging away with the preclear script. However, the post read is going noticably faster (~90 MB/s vs 10 before). Hopefully that's a good sign.

Ok so my preclear has finished on one of my drives, the WD10EACS. I'm not sure what I'm looking for here, but can anyone tell me if this output would indicate a problem?

 

Disk Temperature: 32C, Elapsed Time:  19:40:40
===========================================================================
=                unRAID server Pre-Clear disk /dev/sda
=                       cycle 1 of 1
= Disk Pre-Clear-Read completed                                 DONE
= Step 1 of 10 - Copying zeros to first 2048k bytes             DONE
= Step 2 of 10 - Copying zeros to remainder of disk to clear it DONE
= Step 3 of 10 - Disk is now cleared from MBR onward.           DONE
= Step 4 of 10 - Clearing MBR bytes for partition 2,3 & 4       DONE
= Step 5 of 10 - Clearing MBR code area                         DONE
= Step 6 of 10 - Setting MBR signature bytes                    DONE
= Step 7 of 10 - Setting partition 1 to precleared state        DONE
= Step 8 of 10 - Notifying kernel we changed the partitioning   DONE
= Step 9 of 10 - Creating the /dev/disk/by* entries             DONE
= Step 10 of 10 - Testing if the clear has been successful.     DONE
= Disk Post-Clear-Read completed                                DONE
Disk Temperature: 32C, Elapsed Time:  19:42:09
============================================================================
==
== Disk /dev/sda has been successfully precleared
==
============================================================================
S.M.A.R.T. error count differences detected after pre-clear
note, some 'raw' values may change, but not be an indication of a problem
62,63c62,63
< 192 Power-Off_Retract_Count 0x0032   200   200   000    Old_age   Always       -       17
< 193 Load_Cycle_Count        0x0032   179   179   000    Old_age   Always       -       65996
---
> 192 Power-Off_Retract_Count 0x0032   200   200   000    Old_age   Always       -       18
> 193 Load_Cycle_Count        0x0032   179   179   000    Old_age   Always       -       65997
============================================================================
root@Tower:/boot#

 

As for my WD20EARS drive, it is chugging away with the preclear script. However, the post read is going noticably faster (~90 MB/s vs 10 before). Hopefully that's a good sign.

 

From what I've read by googling that string, it represents an emergency retract of the disk heads when power was lost.

In the same way, the load-cycle-count incremented by one indicating the disk heads were loaded again.  Together they might indicate a loose power connection to the drive where power was lost and then restored.

The system log might have other clues, but the disk is cleared.

Joe L.

Ok, finally finished preclearing the WD20EARS, it took 75 hours!!

===========================================================================
=                unRAID server Pre-Clear disk /dev/sdb
=                       cycle 1 of 1
= Disk Pre-Clear-Read completed                                 DONE
= Step 1 of 10 - Copying zeros to first 2048k bytes             DONE
= Step 2 of 10 - Copying zeros to remainder of disk to clear it DONE
= Step 3 of 10 - Disk is now cleared from MBR onward.           DONE
= Step 4 of 10 - Clearing MBR bytes for partition 2,3 & 4       DONE
= Step 5 of 10 - Clearing MBR code area                         DONE
= Step 6 of 10 - Setting MBR signature bytes                    DONE
= Step 7 of 10 - Setting partition 1 to precleared state        DONE
= Step 8 of 10 - Notifying kernel we changed the partitioning   DONE
= Step 9 of 10 - Creating the /dev/disk/by* entries             DONE
= Step 10 of 10 - Testing if the clear has been successful.     DONE
= Disk Post-Clear-Read completed                                DONE
Disk Temperature: 31C, Elapsed Time:  75:10:26
============================================================================
==
== Disk /dev/sdb has been successfully precleared
==
============================================================================
S.M.A.R.T. error count differences detected after pre-clear
note, some 'raw' values may change, but not be an indication of a problem
58c58
<   7 Seek_Error_Rate         0x002e   200   200   000    Old_age   Always       -       0
---
>   7 Seek_Error_Rate         0x002e   100   253   000    Old_age   Always       -       0
63c63
< 193 Load_Cycle_Count        0x0032   186   186   000    Old_age   Always       -       43015
---
> 193 Load_Cycle_Count        0x0032   186   186   000    Old_age   Always       -       43016
65c65
< 197 Current_Pending_Sector  0x0032   200   200   000    Old_age   Always       -       1
---
> 197 Current_Pending_Sector  0x0032   200   200   000    Old_age   Always       -       0
============================================================================

Should I be worried about the Current Pending Sector or Seek Error Rate changes?

Ok, finally finished preclearing the WD20EARS, it took 75 hours!!

===========================================================================
=                unRAID server Pre-Clear disk /dev/sdb
=                       cycle 1 of 1
= Disk Pre-Clear-Read completed                                 DONE
= Step 1 of 10 - Copying zeros to first 2048k bytes             DONE
= Step 2 of 10 - Copying zeros to remainder of disk to clear it DONE
= Step 3 of 10 - Disk is now cleared from MBR onward.           DONE
= Step 4 of 10 - Clearing MBR bytes for partition 2,3 & 4       DONE
= Step 5 of 10 - Clearing MBR code area                         DONE
= Step 6 of 10 - Setting MBR signature bytes                    DONE
= Step 7 of 10 - Setting partition 1 to precleared state        DONE
= Step 8 of 10 - Notifying kernel we changed the partitioning   DONE
= Step 9 of 10 - Creating the /dev/disk/by* entries             DONE
= Step 10 of 10 - Testing if the clear has been successful.     DONE
= Disk Post-Clear-Read completed                                DONE
Disk Temperature: 31C, Elapsed Time:  75:10:26
============================================================================
==
== Disk /dev/sdb has been successfully precleared
==
============================================================================
S.M.A.R.T. error count differences detected after pre-clear
note, some 'raw' values may change, but not be an indication of a problem
58c58
<   7 Seek_Error_Rate         0x002e   200   200   000    Old_age   Always       -       0
---
>   7 Seek_Error_Rate         0x002e   100   253   000    Old_age   Always       -       0
63c63
< 193 Load_Cycle_Count        0x0032   186   186   000    Old_age   Always       -       43015
---
> 193 Load_Cycle_Count        0x0032   186   186   000    Old_age   Always       -       43016
65c65
< 197 Current_Pending_Sector  0x0032   200   200   000    Old_age   Always       -       1
---
> 197 Current_Pending_Sector  0x0032   200   200   000    Old_age   Always       -       0
============================================================================

 

Should I be worried about the Current Pending Sector or Seek Error Rate changes?

The pending sector is no longer existent, but no re-allocated sectors incremented.  That indicates the original sector was able to be re-written.  The seek-error rate values just changed from their un-used set of values to an initial set of values used once the drive had a few hours on it.  Nothing to worry about there.

I'd worry about the 75 hours.  Its performance is horrible.

I'm thinking that won't actually be an issue. As I described above, I think that was just an issue with the EARS drive. I've read that a lot of people have problems with these drives if they initially use it without a jumper in a Windows machine (as I did) and then try to add a jumper and format it in an unraid server. However, most people have been able to fix the problem by running the preclear script. During the preread, the drive was being a read at an atrocious 10 MB/s, but during the post read it was more in the 70-90 range. Additionally, I've gotten much better transfer rates (30 MB/s vs 1 MB/s before the preclear) as I've been copying data over to it from an external USB drive.

I also used to get a bunch of errors on the drive during boot, but I'm not getting any of that anymore. So my conclusion is that whatever issue the drive was happened, appears to have been rectified by running preclear.

I'd worry about the 75 hours.  Its performance is horrible.

 

I'm thinking that won't actually be an issue. As I described above, I think that was just an issue with the EARS drive. I've read that a lot of people have problems with these drives if they initially use it without a jumper in a Windows machine (as I did) and then try to add a jumper and format it in an unraid server. However, most people have been able to fix the problem by running the preclear script. During the preread, the drive was being a read at an atrocious 10 MB/s, but during the post read it was more in the 70-90 range. Additionally, I've gotten much better transfer rates (30 MB/s vs 1 MB/s before the preclear) as I've been copying data over to it from an external USB drive.

 

I also used to get a bunch of errors on the drive during boot, but I'm not getting any of that anymore. So my conclusion is that whatever issue the drive was happened, appears to have been rectified by running preclear.

Enjoy your server.

Joe L.