[solved] Preclear of multiple drives in RC8 is very slow

[sgibbers17]

I am trying to preclear 5 drives at the same time with RC8 and it is going very slow at less than 1 MB/s once step 2 is reached (writing to remainder of disk).  Step 1 runs at the expected 80 to 100 MB/s.  I swapped to version 4.7 and precleared the same 5 drives at 80 to 100 MB/s on step 2.  Is this a problem others are having?

syslog_11-14-2012.zip

[shat]

My drives all cleared slow too in 8. Who knows. Thought it was just me.

[sgibbers17]

My drives all cleared slow too in 8. Who knows. Thought it was just me.

 

One Drive at a time preclear as expected just when I preclear multiple drives it goes very slow.

[whipdancer]

I precleared 2 drives at the same time (using an esata port multiplier) and speeds dropped at their lowest to the 32 - 35MB/s range.  I'm on 5.0-rc8a.

[sgibbers17]

I precleared 2 drives at the same time (using an esata port multiplier) and speeds dropped at their lowest to the 32 - 35MB/s range.  I'm on 5.0-rc8a.

 

I am using a SUPERMICRO AOC-SASLP-MV8.  I think that my controller may not be fully working under the new v5-RC8.  This controller does work under v4.7.  I'm going to place my drives on the MB and see what happens.

[sgibbers17]

I have uploaded a syslog and I think I was remembering incorrectly when I said that one drive will preclear fine one drive will also preclear very slowly at about 400 to 500 KB/s.  But preclearing on version 4.7 works just fine.  I need to use version 5 because I have 3TB drives that I want to use.

syslog_11-14-2012.zip

[atom]

Hi all,

 

I think I may be experiencing the same issue. I'm running 5.0-rc8a on an Asus C60M1-I. Currently preclearing 2 3 TB drives (WD30EZRX) at the same time. The console shows ~65MB/s but unMenu shows ~28MB/s. Going on 46h and have not finished the 1st cycle yet. I need v5 for the 3TB drives so I guess I'm stuck.

 

================================================================== 1.13
=                unRAID server Pre-Clear disk /dev/sdb
=               cycle 1 of 3, partition start on sector 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 - Verifying if the MBR is cleared.              DONE
= Post-Read in progress: 94% complete. 
(  2,845,946,880,000  of  3,000,592,982,016  bytes read ) 65.9 MB/s
Disk Temperature: 29C, Elapsed Time:  46:08:16

[sgibbers17]

Here is a copy of my syslog running 4.7 working correctly preclearing 1.5TB drives.

syslog_v4.7.zip

[sheppp]

Count me as another of those experiencing painfully slow preclear speeds while preclearing multiple drives under RC8 (four 4tb Hitachi drives on one SUPERMICRO AOC-SASLP-MV8, four more on another SASLP-MV8, and a ninth 4tb WD Black on a SATA III port on a SUPERMICRO MBD-X9SCM-F-O MB). As mentioned previously, step 1 ran at normally expected speeds; step 2, however, is another matter. Although I can see that bytes are being slooowwwwwly written to the drives, I have been stuck at zero % for approx. 18 hours now (example of my 4tb WD Black: Wrote  29,307,699,200  bytes out of  4,000,787,030,016  bytes (0% Done)).

 

Initially I attempted to preclear 8 of the drives on a sole SASLP-MV8. But the whole thing crashed after a couple of hours. I posted my syslog, but have heard nothing back re whether this result should be expected while attempting such a thing on a single card.

 

Obviously, there is a problem somewhere. But, I have no idea where it is or how to fix it.

[twolf]

As a X9SCM-F user, you might want to check out this thread: http://lime-technology.com/forum/index.php?topic=22675.0

 

[sheppp]

Thanks. I've had other problems with this new X9SCM-F board. One of the NIC ports doesn't work. SuperMicro tech support has been very responsive but can't seem to resolve the issue. I also had - what I thought - were problems booting up only 2 of the 3 AOC-SASLP-MV8s with it. After multiple hours of tech support assistance, it was finally determined that although only 2 of them appeared to boot up, once I fired up UnRAID, I could see drives on all three. According to tech support: "The BIOS needs a certain amount of option ROM in order to boot up therefore it has to limit the amount of option ROM that can be used by add-on cards. Once you go into the OS, you will be able to see all 3 cards." This proved to be correct.

 

This is my second UnRAID box that I have built (the first being with an ASUS board without IPMI - which is what I wanted with this build). I got the X9SCM-F and the SASLP-MV8s based on the feedback in this forum. Unfortunately, I either had blinders on or I failed to see problems associated with them. If anybody has any suggestions for a MB with IPMI that will take an Intel Xeon E3-1230 cpu, I'm all ears. Also, if the SASLP-MV8s are problematic, I would also like to find a solution to take the place of the the 3 SASLP-MV8s that I currently have (22 SATA ports to max out UnRAID). I've seen people using an IBM M1015 card with an expander, but I don't have an extra box to reflash it, they are getting more difficult to find, and they seem a bit expensive now compared to others out there.

 

Any assistance will be greatly appreciated.

[sgibbers17]

Once I updated to RC9 I was able to preclear three 3TB drives at once at about 100MB/s.