Recommended Posts

Even more inportant since the drive is new. Do you truely trust your new drive? Its not really been in service so by preclearing its going to test your new drive in a way that it would never run. Several people in the past have said their brand new drives right out of the box have failed preclear and had to send their drives back for new ones.

 

After I've successfuly ran preclear on all my drives I know they are going to serve me well and have since I've installed them. Cheap insurance

Link to comment

Even more inportant since the drive is new. Do you truely trust your new drive? Its not really been in service so by preclearing its going to test your new drive in a way that it would never run. Several people in the past have said their brand new drives right out of the box have failed preclear and had to send their drives back for new ones.

 

After I've successfuly ran preclear on all my drives I know they are going to serve me well and have since I've installed them. Cheap insurance

Based on feedback in the pre-clear thread, somewhere between 1 out of 5 and 1 out of 10 drives seems to be defective out of the box.  How valuable is your data to you?    Do you think you're lucky? 

 

Serious users run each disk through 3 pre-clear cycles before using it in their server.  (preclear_disk does not just clear the disk, but exercises it in a way intended to expose any physical or electrical weakness.)  Yes, it takes time... but is a LOT easier replacing the drive before you put your data on it than after.

 

Joe L.

Link to comment

Even more inportant since the drive is new. Do you truely trust your new drive? Its not really been in service so by preclearing its going to test your new drive in a way that it would never run. Several people in the past have said their brand new drives right out of the box have failed preclear and had to send their drives back for new ones.

 

After I've successfuly ran preclear on all my drives I know they are going to serve me well and have since I've installed them. Cheap insurance

Based on feedback in the pre-clear thread, somewhere between 1 out of 5 and 1 out of 10 drives seems to be defective out of the box.   How valuable is your data to you?    Do you think you're lucky? 

 

Serious users run each disk through 3 pre-clear cycles before using it in their server.  (preclear_disk does not just clear the disk, but exercises it in a way intended to expose any physical or electrical weakness.)   Yes, it takes time... but is a LOT easier replacing the drive before you put your data on it than after.

 

Joe L.

 

I've done preclears now on about 7 of the WD 2TB green drives, these would take about 48 to 49 hours to do two preclear cycles.  On the sixth drive the preclear took over 60 hours to do the first cycle and I gave up when it had got past 100 hours and the second pass was not yet done.  I took that drive back and got a replacement which only took the normal 48 hours to do two preclear passes.  For some reason the writes on this bad drive were a lot slower, but the reads seemed to be about the normal speed.  I'm glad I didn't put that drive into my array as it would have slowed writes down a lot.  The SMART reports did not show anything particularly bad about it, just it had a much higher load cycle count than the good drives did (like about 4000 versus 50 after the preclearing).

 

Stephen

 

Link to comment
I've done preclears now on about 7 of the WD 2TB green drives, these would take about 48 to 49 hours to do two preclear cycles.  On the sixth drive the preclear took over 60 hours to do the first cycle and I gave up when it had got past 100 hours and the second pass was not yet done.  I took that drive back and got a replacement which only took the normal 48 hours to do two preclear passes.  For some reason the writes on this bad drive were a lot slower, but the reads seemed to be about the normal speed.  I'm glad I didn't put that drive into my array as it would have slowed writes down a lot.  The SMART reports did not show anything particularly bad about it, just it had a much higher load cycle count than the good drives did (like about 4000 versus 50 after the preclearing).

 

Stephen

 

 

How do you describe the issue and go through the hoops of making sure it's actually needing RMA to the manufacturer? I've never sent back a hard drive.

Link to comment

Just say that it fails SMART tests.  Newegg and other retailers will accept that reason at face-value.  Drive manufacturers may make you run their utilities (which are just their versions of SMART) and give them the output code.  Also, always insist that they cover your shipping fees, as they often try to make you pay it.

Link to comment

How do you describe the issue and go through the hoops of making sure it's actually needing RMA to the manufacturer? I've never sent back a hard drive.

 

I get all my drives from a local dealer, I just told them it failed the raid array's burn in tests (as it was too slow) and they exchanged it for a new one on the spot.

 

Stephen

 

Link to comment

I just received my two 2TB WD20EARS.  I'm running them through the Western Digital Data Lifeguard Diagnostics long smart test while I build my unraid server.  So far, one passed and the other is being tested.  (Takes about 5 hours.)  Is the Lifeguard Diagnostics test revealing of disk issues?

 

I am going to preclear them starting tomorrow.  If I wanted to let them go through the test three times, what command line parameters would I use if my EARS disks are jumpered (7-8)?

 

Also, I know I can preclear multiple disks at the same time but can I use preclear while the server is serving files or does it need to be stopped?

 

Thanks for the help.

 

jsdds

Link to comment

I just received my two 2TB WD20EARS.  I'm running them through the Western Digital Data Lifeguard Diagnostics long smart test while I build my unraid server.  So far, one passed and the other is being tested.  (Takes about 5 hours.)  Is the Lifeguard Diagnostics test revealing of disk issues?

 

I am going to preclear them starting tomorrow.  If I wanted to let them go through the test three times, what command line parameters would I use if my EARS disks are jumpered (7-8)?

 

Also, I know I can preclear multiple disks at the same time but can I use preclear while the server is serving files or does it need to be stopped?

 

Thanks for the help.

 

jsdds

 

If you are jumpering the drives, then you want to use the "-a" paramter to preclear (that is the default).  It will cause the drive to align the first partition at sector 63.  A bit confusing, but the jumper causes the drive to interpret sector 63 as sector 64 internally, so this will give optimal performance.  The other option would be to remove the jumper and use the "-A" parameter.  This will align the first partition at sector 64, and prevent the drive from doing its little sector kludge.  If you go with the second option, however, you'll have to run unRAID 4.7b1 (or the new 5.0b3 which is not out yet).

 

The command to run the drive through three cycles, with the jumper, is ...

 

preclear_disk.sh -a -c 3 /dev/sdX

 

where sdX is the device of your disk.

 

The command to run the drive through three cycles, with NO jumper, is ...

 

preclear_disk.sh -A -c 3 /dev/sdX

 

You might want to give THIS THREAD a good read before running preclear.  And download the latest and greatest version of preclear from the bottom of that thread as well.

 

Suggest you run the commands on the server console itself (you can run two at the same time because the console enables up to 6 sessions by pressing Alt-F1 - Alt-F6).  Or you can also use the "screen" addon to run multiple sessions.  Just be careful not to start preclear in a simple telnet window, because if you do, when the window closes, the preclear will terminate.

 

Good luck.

Link to comment

As long as you do not close the telnet windows, it should be fine.

 

As I discovered today, there is a limit on concurrent telnet sessions - can anyone confirm that opening a 9th telnet session you get an "UNAUTHORIZED ROOT LOGIN" in syslog?

 

 

That is because there are only 8 devices in the linux kernel defined to log in on.

ls -l /dev/ptyp*

crw-rw---- 1 root tty  2, 0 Jan 21 11:21 /dev/ptyp0

crw-rw---- 1 root tty  2, 1 Jan 21 11:21 /dev/ptyp1

crw-rw---- 1 root tty  2, 2 Jan 21 11:21 /dev/ptyp2

crw-rw---- 1 root tty  2, 3 Jan 21 11:21 /dev/ptyp3

crw-rw---- 1 root tty  2, 4 Jan 21 11:21 /dev/ptyp4

crw-rw---- 1 root tty  2, 5 Jan 21 11:21 /dev/ptyp5

crw-rw---- 1 root tty  2, 6 Jan 21 11:21 /dev/ptyp6

crw-rw---- 1 root tty  2, 7 Jan 21 11:21 /dev/ptyp7

Link to comment

As long as you do not close the telnet windows, it should be fine.

 

This is true, but it can take 30 hours or more for a preclear - more if you are running multiple cycles.  If anything happens and you have to reboot your workstation, or you accidently close your telnet session, you have to start over. 

 

I DO NOT recommend running preclear in a telnet window without using screen.  Run from the console (i.e., keyboard/display hooked to the unRAID server) or using screen via a telnet window.

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.