Hi All,

 

I've been running 4.7 for a long time.  In anticipation of migrating to latest 5.0 rc release (plus forum samba fix) I picked up a couple of WD 3TB drives.  I did a /boot/preclear_disk.sh -A sdc under screen for the first 3TB drive.  They are both hanging off my AOC-SASLP-MV8 (latest .21 FW).  All seemed to be going well until the preclear script stalled at this stage (it hasn't moved in a couple of hours):

 

================================================================== 1.13

=                unRAID server Pre-Clear disk /dev/sdc

=              cycle 1 of 1, 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: 90% complete.

(  2,702,827,008,000  of  3,000,592,982,016  bytes read ) 75.3 MB/s

Disk Temperature: 26C, Elapsed Time:  30:03:35

 

I can still access the machine via telnet, but Web OI is non-responsive and trying to read any files on the raid drives via samba usually fails.  Some files failed half an hour ago but now are readable??  No zombies and nothing unusual under top other than most of the 4GB RAM is used and nothing related to the pre-clear is hogging anything as I'd expect (with a working system).

 

Bootup after successful parity check:

http://pastebin.ca/2299213

 

First error shortly after preclear had started:

http://pastebin.ca/2299215

 

SAS error spewage every 10s a few hours later:

http://pastebin.ca/2299216

 

Throw in some reiserfs unhappiness for good measure:

http://pastebin.ca/2299218

 

I can still telnet into the machine.  The web interface is inaccessible.  What should I do?  Other than "Don't Panic"

 

1) Kill the pre-clear that was "nearly" finished?

2) Do a reiserfs check?

3) If 2 is ok, try pre-clearing again?

 

If I try another pre-clear I'll not use the box for anything other than running screen while it does it's thing.  I did some reading and writing during the pre-clear this time.  Since it wasn't to the drive being cleared I figured it'd be OK, just maybe slower than usual.

 

Thanks for any help!

 

P.S. Full 25MB syslog is attached as a .7z file to squeak under the forum max size.  Renamed to .zip in case .7z was disallowed.

syslog.zip

The web OI eventually responded although all the SAS drives show as 0 deg C temp.  Sure enough I can samba the non-SAS drive files.  So it looks like I had a fatal SAS error during the pre-clear.  Can I move existing full drives onto the SAS card and do the pre-clears via motherboard SATA ports in the hopes that  will have a better chance of working and all SAS errors will go away with 5.0? 

4.7 OS does not support 3TB drives.

4.7 OS does not support 3TB drives.

 

Thanks Joe.  For some reason I was under the impression that the preclear script could be used on 3TB drives in 4.7 but you couldn't add them to the array, until upgrading to 5.

I thought I'd seen this somewhere on here: 

 

Can I preclear a 3TB drive under unRAID 4.7 so it will be ready to use in its fullest under unRAID v5.0 latest beta?

yes.    Just make sure you use the latest version of the preclear script (1.13 is the newest at this time)

 

Joe L.

 

This is exactly what I'd like to do.

4.7 OS does not support 3TB drives.

 

Thanks Joe.  For some reason I was under the impression that the preclear script could be used on 3TB drives in 4.7 but you couldn't add them to the array, until upgrading to 5.

It will entirely depend upon your disk controllers and their affiliated drivers.  They are still involved.  If the driver cannot access disk blocks > 2.2TB, then it will not work.  No idea if that is your issue, as all kinds of stuff seems to be happening on your server.

 

Joe L.

Yeah, I figured the SASLP being so popular would be OK.  What about my idea of re-arranging the drives so the 3TB one's are on the MB?  Is that doable or is preclearing 3TB drives on 4.7 just plain risky?  Thanks!

P.S. It looks like the preclear did quite a bit before failing.  Does it only access > 2.2TB after step10?

You can re-arrange drives. You just have to re-assign them to the proper slots when you boot. You should have better luck using the motherboard.

 

FYI, once you're on the newer 5 series you should be able to re-arrange drives without having to re-assign anything since it uses the drive ID and not the port for the assignments.

Just an update for people form the future. 

 

I re-arranged my drives so that the 2 new 3TB drives were on the MB (6 ports) rather than the SASLP card (8 ports),  along with parity and my only non-full drive.  I keep > 20GB free on all my drives.    Re-booted and re-assigned disks based on original device list screenshot, and array came up fine.  Ran the ReiserFS tests on all the drives and only 1 of the 9 needed a fix and it fixed no problem (no lost+found).  Finally re-ran preclear on both 3TB drives with 1.13 version of script and unraid 4.7 and they both finished with no issues.  Looks like my SASLP might have some issues on 4.7 when stressed.

 

And all just in time to switch to 5.0 if this rc9 (edit: make that rc9a, lol!) looks ok! 

 

Thanks