Burn-in for new drives -- not sure if I am running the right application


Recommended Posts

I've started something, but I have to admit, I may be running the wrong program for what I wanted.

 

I downloaded to my flash drive "preclear".   I thought it was a program that tests the hard-drive and prepares it for use in the array, however I just read a post that makes me thinks it just prepares the drive and doesn't test it.

 

3 Questions:

 

1.  Am I running the program for the wrong reasons?

2.  If Yes to 1., can I safely stop it?  (EDIT: Too late for this one as I accidentally rebooted the machine)

3.  If Yes to 1., what program should I be running.

 

I also think I am missing something else on my USB Flash drive as when I run preclear, I get this error at the top:

Pre-Clear unRAID Disk

########################################################################

smartctl: error while loading shared libraries: libstdc++.so.6: cannot open shared object file: No such file or directory

 

 

I think I am getting overwhelmed at times with the software part of getting my new server going.   I honestly have at least 12 tabs open in my browser as each time I look for one answer, I get diverted to multiple others threads as I need more information or I need to know how to do the basis steps.   Not having Linux skills, beyond knowing the command "ls" slows down the process a lot, but... I will continue the battles and I will prevail!  :)

Link to comment
1.  Am I running the program for the wrong reasons?

No, you're doing it for the correct reasons.

The error you get is because there is a missing C++ library for smartctl.

This is a known issue and a search on the board will reveal the packages needed to install for smartctl to work.

Just let the pre-clear run till the end. If you do not, then unRAID and emhttp will do it when you add it to the array.

Link to comment

I also think I am missing something else on my USB Flash drive as when I run preclear, I get this error at the top:

Pre-Clear unRAID Disk

########################################################################

smartctl: error while loading shared libraries: libstdc++.so.6: cannot open shared object file: No such file or directory

You can load the missing library now, even though the preclear script is running.  It will help with the final report.

 

Paste the line below at a command prompt  (It is a long line, ending at the closing paren ")" ).

[pre]

(mkdir /boot/packages;cd /boot/packages;wget http://slackware.cs.utah.edu/pub/slackware/slackware-12.0/slackware/a/cxxlibs-6.0.8-i486-4.tgz;'>http://slackware.cs.utah.edu/pub/slackware/slackware-12.0/slackware/a/cxxlibs-6.0.8-i486-4.tgz; installpkg cxxlibs-6.0.8-i486-4.tgz;)[/pre]

 

Or, type the 4 commands in turn

 

mkdir /boot/packages

cd /boot/packages

wget http://slackware.cs.utah.edu/pub/slackware/slackware-12.0/slackware/a/cxxlibs-6.0.8-i486-4.tgz

installpkg /boot/packages/cxxlibs-6.0.8-i486-4.tgz

 

Either will create a /boot/packages directory, change directory to it, download the missing library, and then install it.

 

If you reboot you'll need to re-run the last line to re-install the package.  No need to download it again, as you already did that once.

 

Joe L.

Link to comment

Thanks guys.  I have my two new drives running through the process now.  My Seagate is flying through quicker than the WD Green ( as expected ).

 

Drive temperatures seem good, which tells me I have enough cooling ( CoolMaster 590 Case, stock case fan, 2 top mount fans and 3 drive cage fans )

 

After 5 hours, the Seagate is at 31C and the WD is at 27C . 

Link to comment

After 8 hours I am 63% done, Stage 2, on my Seagate and 48% on my WD.  What I was suprised to see is the WD is now out performing the Seagate doing the 'Copying zero's to the remainder of the disk' .    Seagates running at 74 MB/s and WD at 91.3 MB/s .

 

Temperatures are holding steady at 31C and 28C.

Link to comment

I've started something, but I have to admit, I may be running the wrong program for what I wanted.

 

I downloaded to my flash drive "preclear".   I thought it was a program that tests the hard-drive and prepares it for use in the array, however I just read a post that makes me thinks it just prepares the drive and doesn't test it.

Perhaps I can clairify...

 

The original reason I wrote the preclear_disk.sh script was to clear them to all zeros except for the special signature pattern that unRAID would use to recognize a drive was pre-cleared and not go through a lengthy clearing of the drive while your array was off-line.  The goal was to minimize the off-line time of the array.  (Once your family gets used to watching movies, etc from the server, you'll understand.  They will not want any downtime... and the thought of a 4 to 6 or 8 hour outage might put you in the dog-house)

 

Once that was working, a member suggested it would be really great use the same routine to burn-in a drive and also to check the built-in diagnostics that the SMART firmware on the disks for any signs of errors.

 

it turned out that SMART will flag a sector as needing re-allocation if it cannot be read.  That led me to add a complete pre-read of the entire drive.  If a bad sector existed, the subsequent "write" of a zero to the bad spot on the disk would allow the SMART firmware to move (re-allocate) the bad sector to one in its reserve of spare sectors. 

 

The post-read was to subsequently flag any additional sectors that might show up as bad even after being re-allocated.  Up until now, I verified the fiirst 512 bytes I wrote on the disk were as expected, but let the SMART firmware on the disk deal with any other issues.

 

Recently (yesterday) a posted a new version of the pre-clear script that in the post-read phase actually verified it was reading all zeros.  Robbie Ferguson does not know it, but his experience with a defective disk gave me the idea to add this check.  It is not perfect, as we can still have a defective disk where a "1" cannot be written, but a "0" can, and the test would sill pass.  But, it is a nice addition, even if it only helps one person out of a hundred to learn their drive is un-reliable.

 

The major method of identification of bad drives is the SMART firmware on the drive, the results and status read by the "smartctl" program in Linux.

 

Now that you have installed the missing c++ support library it will work. (one indication it is working is you now see temperature readings from the drives... they are part of the smart status returned by the smartctl program.)

 

So... all of the versions helped to burn in a drive, even the first one that only wrote zeros and did no pre-post read.

All of the versions used the SMART firmware on the disk to test its performance.

The newest version just adds one more verification to those already being performed.  just in case it helps somebody.

 

As far as your interesting performance differences... One of your drives seems to be better at "read" performance, the other at "write" performance.  Each taking the lead in different portions of the preclear_disk.sh script. 

 

Joe L.

Link to comment

Thanks Joe!  It definitely is a great idea.  Considering the industries continual pressure to produce "bigger" , "faster" and "cheaper" , one has to wonder if the area of quality control is reduced , so being able to test the drive , as best we can ourselves, is a definite bonus.

 

I figured the Seagate drive would out perform the WD in all areas ( due the drive speed difference ), but the WD is holding its ground against it.  The Seagate will finish all 10 steps of the test before the WD.  Both are at Stage 10, after 14 hours, with the Seagate at 36% and the WD at 25% , Seagate reading at 106 MB/s and the WD at 85.9 MB/s and temperatures of 31C and 26C respectively.

 

What is the average number of cycles, one would suggest, to run the application?

 

Side Note:  If you get the CoolMaster 590 Case, you won't need a night-light in the room you put it.  I just wanted into my den , where I have it stored , and the room is well lit.  :)

Link to comment

Thanks Joe!   It definitely is a great idea.   Considering the industries continual pressure to produce "bigger" , "faster" and "cheaper" , one has to wonder if the area of quality control is reduced , so being able to test the drive , as best we can ourselves, is a definite bonus.

Yes, local testing has to be far longer than any tests done at the factory.  They can't possibly be doing much more than the "short" smart tests on the drives while they are on the production line.  it would simply take forever to get anything off the factory floor.

I figured the Seagate drive would out perform the WD in all areas ( due the drive speed difference ), but the WD is holding its ground against it.   The Seagate will finish all 10 steps of the test before the WD.   Both are at Stage 10, after 14 hours, with the Seagate at 36% and the WD at 25% , Seagate reading at 106 MB/s and the WD at 85.9 MB/s and temperatures of 31C and 26C respectively.

Lots of factors dictate overall speed.  Are both disks on the same disk controller?  No  matter what, any given disk must spin its platter a full revolution between the "read" of the existing contents of a sector and the "write" of its updated contents.  Basic physics says the 7200 RPM drive will be able to perform the two operations in sequence faster than the 5400 RPM drive. 

 

In the preclear_script, the memory buffer in the disks used for look-ahead will make a huge difference.  I do not read/then immediately write the same sector.  The read and write phases are usually separated by hours.

What is the average number of cycles, one would suggest, to run the application?

I don't know of an average.  I would guess most people just run 1.  I personally like two or three, but I run them one at a time so I can evaluate the results after each pass.  I only remember one person who reported running 20 cycles of a large disk, but I'm sure others have done the same.  It all depends on how urgently you need the extra space for your files and want to get the disk added to your array.

 

Side Note:  If you get the CoolMaster 590 Case, you won't need a night-light in the room you put it.  I just wanted into my den , where I have it stored , and the room is well lit.  :)

I'll stay away from it then as a consideration for use in my theater... every pilot light makes a HUGE difference when you have a totally light controlled room.  Ruins the contrast ratio.... and ruins the effect when the director "fades to black"

 

Joe L.

Link to comment

My first drive has completed the 1st burn-in cycle of my Seagate 1.5TB drive at just under 21 hours.    I'm using putty , so this is as much as I am able to cut/paste.  The part I high-lighted in red is a good sign.  Is their anything else from the output that I should be concerned about ?  Thanks in advance (again) :)

 

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

= Post-Read in progress: 99% complete.

(  1,500,291,072,000  of  1,500,300,828,160  bytes read ) 61.7 MB/s

Disk Temperature: 30C, Elapsed Time:  20:38:57

===========================================================================

=                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: 30C, Elapsed Time:  20:39:54

============================================================================

==

== 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

54c54

<  1 Raw_Read_Error_Rate    0x000f  119  100  006    Pre-fail  Always      -      224929533

---

>  1 Raw_Read_Error_Rate    0x000f  117  099  006    Pre-fail  Always      -      151837309

58c58

<  7 Seek_Error_Rate        0x000f  100  253  030    Pre-fail  Always      -      1133

---

>  7 Seek_Error_Rate        0x000f  100  253  030    Pre-fail  Always      -      535283

64,66c64,66

< 189 High_Fly_Writes        0x003a  100  100  000    Old_age  Always      -      0

< 190 Airflow_Temperature_Cel 0x0022  070  068  045    Old_age  Always      -      30 (Lifetime Min/Max 29/30)

< 195 Hardware_ECC_Recovered  0x001a  050  040  000    Old_age  Always     

---

> 189 High_Fly_Writes        0x003a  083  083  000    Old_age  Always      -      17

> 190 Airflow_Temperature_Cel 0x0022  070  068  045    Old_age  Always      -      30 (Lifetime Min/Max 29/32)

> 195 Hardware_ECC_Recovered  0x001a  055  040  000    Old_age  Always     

70,72c70,72

< 240 Head_Flying_Hours      0x0000  100  253  000    Old_age  Offline      -      166236709191681

< 241 Unknown_Attribute      0x0000  100  253  000    Old_age  Offline      -      2033

< 242 Unknown_Attribute      0x0000  100  253  000    Old_age  Offline      -      333774446

---

> 240 Head_Flying_Hours      0x0000  100  253  000    Old_age  Offline      -      72782515798038

> 241 Unknown_Attribute      0x0000  100  253  000    Old_age  Offline      -      1819551215

> 242 Unknown_Attribute      0x0000  100  253  000    Old_age  Offline      -      292686191

============================================================================

root@Tower:/boot#

 

Link to comment

My first drive has completed the 1st burn-in cycle of my Seagate 1.5TB drive at just under 21 hours.    I'm using putty , so this is as much as I am able to cut/paste.   The part I high-lighted in red is a good sign.  Is their anything else from the output that I should be concerned about ?  Thanks in advance (again) :)

Looks pretty good to me.

 

Joe L.

Link to comment

If the creator of the application says its good, then that's good enough for me!  Smiley

 

The WD has now completed.  Its results have less "stuff" after the declaration it was successful:

 

 

===========================================================================

=                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

= Post-Read in progress: 99% complete.

(  1,500,291,072,000  of  1,500,301,910,016  bytes read ) 46.1 MB/s

Disk Temperature: 28C, Elapsed Time:  22:43:40

===========================================================================

=                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: 29C, Elapsed Time:  22:44:49

============================================================================

==

== 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

63c63

< 193 Load_Cycle_Count        0x0032   200   200   000    Old_age   Always                          -       25

---

> 193 Load_Cycle_Count        0x0032   200   200   000    Old_age   Always                          -       27

============================================================================

root@Tower:/boot#

 

 

 

Link to comment

ok, so I currently have one drive running through the tower.  Presuming I can telnet into the tower from my PC (hint: detailed instructions from Vista x64 would be great!), I would have to run 2 more sessions to do the other 2 disks.  In the meantime, I'll start poking around the forum for info on telinet'ing.

Link to comment

Download "putty"

 

Open it, click on the little button beside "telnet" , in the box for address, just put "tower"

 

Login as: root

No password

 

You need to find the ID of the drive you are pre-clearing first, provided by Joe L in another thread:

 

If you are unsure of the device for your disk, type:

ls -l /dev/disk/by-id and look for the device affiliated with your disk's model/serial-number.

 

To use pre-clear, unzip onto your flash drive, then log in via telnet, cd to the folder you unzipped it into, and then type:

preclear_disk.sh /dev/aaa

 

Where "/dev/aaa" =  /dev/hda, /dev/hdb, /dev/hdc, /dev/hdd ... or /dev/sda, /dev/sdb, /dev/sdc, etc.

 

 

Clear as mud?  :)

 

 

Link to comment

If something was wrong with the drive, does it say something like:

 

Disk /dev/sdb has NOT been successfully precleared

That is almost exactly what it says.

Also, when you run your 2nd and 3rd cycles, do you put any switches after the file name ( ie. are their steps you don't need to repeat that can speed it up )?

No, you really want all those steps.

You can see the available options by typing

preclear_disk.sh -?

although it will scroll off the screen, so you'll need to scroll back up to see them.

about the only one you might want is

preclear_disk.sh -c 3 /dev/sdX

 

to have it do three cycles in sequence.

The options always are entered before the name of the device being cleared.

Link to comment

Download "putty"

 

Open it, click on the little button beside "telnet" , in the box for address, just put "tower"

 

Login as: root

No password

 

You need to find the ID of the drive you are pre-clearing first, provided by Joe L in another thread:

 

If you are unsure of the device for your disk, type:

ls -l /dev/disk/by-id and look for the device affiliated with your disk's model/serial-number.

 

To use pre-clear, unzip onto your flash drive, then log in via telnet, cd to the folder you unzipped it into, and then type:

preclear_disk.sh /dev/aaa

 

Where "/dev/aaa" =  /dev/hda, /dev/hdb, /dev/hdc, /dev/hdd ... or /dev/sda, /dev/sdb, /dev/sdc, etc.

 

 

Clear as mud?  :)

 

 

 

Thanks for the all the info Spinbot.  I had no clue what the names of my disks were until I installed UnMenu (thanks Joe L.!).  Sounds like once I use telnet then it's as if I'm working directly from the server UI.  I think I'll go ahead and start the other 2 disks ASAP.  I appreciate your help!

Link to comment

I ran the test again and this time, the results didn't come back as good on the Seagate drive.   The only change I have made is that I ripped out one of my 1GB sticks of RAM.    How should I interpret these?

 

 

===========================================================================

=                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

= Post-Read in progress: 99% complete.

(  1,500,291,072,000  of  1,500,300,828,160  bytes read ) 60.8 MB/s

Disk Temperature: 31C, Elapsed Time:  20:05:38

===========================================================================

=                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: 31C, Elapsed Time:  20:06:34

============================================================================

==

== Disk /dev/sda has NOT been precleared successfully

== skip=1800 count=200 returned 45060 instead of 00000 skip=4400 count=200 returned 46085 instead of 00000 skip=6600 count=200 returned 54273 instead of 00000 skip=9600 count=200 returned 00005 instead of 00000 skip=12000 count=200 returned 45056 instead of 00000 skip=20000 count=200 returned 12289 instead of 00000 skip=20200 count=200 returned 46081 instead of 00000 skip=41000 count=200 returned 61445 instead of 00000 skip=46400 count=200 returned 24584 instead of 00000 skip=48400 count=200 returned 32773 instead of 00000

============================================================================

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

54c54

<   1 Raw_Read_Error_Rate     0x000f   117   099   006    Pre-fail  Always       -       151884615

---

>   1 Raw_Read_Error_Rate     0x000f   114   099   006    Pre-fail  Always       -       80042350

58c58

<   7 Seek_Error_Rate         0x000f   100   253   030    Pre-fail  Always       -       557279

---

>   7 Seek_Error_Rate         0x000f   060   060   030    Pre-fail  Always       -       1099614

64,66c64,66

< 189 High_Fly_Writes         0x003a   083   083   000    Old_age   Always       -       17

< 190 Airflow_Temperature_Cel 0x0022   069   068   045    Old_age   Always       -       31 (Lifetime Min/Max 31/31)

< 195 Hardware_ECC_Recovered  0x001a   045   040   000    Old_age   Always     

---

> 189 High_Fly_Writes         0x003a   073   073   000    Old_age   Always       -       27

> 190 Airflow_Temperature_Cel 0x0022   069   068   045    Old_age   Always       -       31 (Lifetime Min/Max 30/32)

> 195 Hardware_ECC_Recovered  0x001a   056   040   000    Old_age   Always     

70,72c70,72

< 240 Head_Flying_Hours       0x0000   100   253   000    Old_age   Offline      -       32182189948954

< 241 Unknown_Attribute       0x0000   100   253   000    Old_age   Offline      -       1819551215

< 242 Unknown_Attribute       0x0000   100   253   000    Old_age   Offline      -       292689871

---

> 240 Head_Flying_Hours       0x0000   100   253   000    Old_age   Offline      -       52067888529454

> 241 Unknown_Attribute       0x0000   100   253   000    Old_age   Offline      -       1213618071

> 242 Unknown_Attribute       0x0000   100   253   000    Old_age   Offline      -       16484769

============================================================================

root@Tower:/boot#

 

Link to comment

This is BAD.

 

Disk /dev/sda has NOT been precleared successfully

== skip=1800 count=200 returned 45060 instead of 00000 skip=4400 count=200 returned 46085 instead of 00000 skip=6600 count=200 returned 54273 instead of 00000 skip=9600 count=200 returned 00005 instead of 00000 skip=12000 count=200 returned 45056 instead of 00000 skip=20000 count=200 returned 12289 instead of 00000 skip=20200 count=200 returned 46081 instead of 00000 skip=41000 count=200 returned 61445 instead of 00000 skip=46400 count=200 returned 24584 instead of 00000 skip=48400 count=200 returned 32773 instead of 00000

 

Now, it could be the disk, or bad memory, or a bad disk controller,  but it appears as if the drive was not zeroed. 

(It read back what it thought should have been all zeros, and they were not)

 

You can try another pass at it, and see if the spots returning non-zeros move... (or go away)  But right now, that disk does not appear to be cleared.  If you suspect the RAM, then this is probably confirming it is still bad.

 

Joe L.

 

 

 

Link to comment

I ran the test again and this time, the results didn't come back as good on the Seagate drive.   The only change I have made is that I ripped out one of my 1GB sticks of RAM.    How should I interpret these?

 

 

This is BAD.

 

Now, it could be the disk, or bad memory, or a bad disk controller,  but it appears as if the drive was not zeroed.

(It read back what it thought should have been all zeros, and they were not)

 

You can try another pass at it, and see if the spots returning non-zeros move... (or go away)  But right now, that disk does not appear to be cleared.  If you suspect the RAM, then this is probably confirming it is still bad.

 

Interpret this that the bad memory is scrambling data before it gets written to disk.

I mentioned this could happen with bad memory.

http://lime-technology.com/forum/index.php?topic=4404.msg39536#msg39536

Link to comment

Stop discouraging me!  :)

 

I'm getting ready for the "close your eyes and the problem doesn't exist approach" pretty soon.     The more testing I do, the further I get to my goal.

 

I'm defeated tonight.   Going to give this a rest for a night and return to the battle tomorrow.

 

There is no battle.

Get reliable memory. If you are not passing memtest 100% of the time nothing can be relied upon.

Link to comment

Stop discouraging me!  :)

 

I'm getting ready for the "close your eyes and the problem doesn't exist approach" pretty soon.     The more testing I do, the further I get to my goal.

 

I'm defeated tonight.   Going to give this a rest for a night and return to the battle tomorrow.

Get a good night's rest.  If you were to put your array to work on real data now, the odds of it doing successive successful parity checks are very tiny... and the parity protection of your precious data non-existent.

 

You do have the honor of showing me how I can improve the pre-clear script a tiny bit more.  I think it would benefit with a tiny addition of printing the block size used in the final verify step.  I'm sure it is in the syslog, but it should be with the failure results too.   You are the first with a printout showing a failed verify of the zeroed drive.  

 

As I said, odds are the drive is good, but something else is not letting it read properly from the disk.  You'll learn more as you test other disks.  (And I'll learn too)

 

Joe L.

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.