Preclear.sh results - Questions about your results? Post them here.

[prostuff1]

Those results are basically identical to the ones you posted above.  My same suggestion stands for the above drive.

[Joe L.]

Those results are basically identical to the ones you posted above.  My same suggestion stands for the above drive.

If that drive was in the same "windows" server, it makes me think that there was something in it preventing proper "writes" to the drives... perhaps a marginal power supply?  Something common which preventing the writes from occurring properly on the drives.  Again all the sectors marked as un-readable were able to be re-written in place, and not actually re-allocated at all.

[ricsouthcott]

yeah i had all the drives hooked up to the a windows  server  and then i bought unraid

[Joe L.]

yeah i had all the drives hooked up to the same server  basicially i have used all the drives and bought unraid

Is it the same motherboard and power supply?

[ricsouthcott]

yeah i had all the drives hooked up to the same server  basicially i have used all the drives and bought unraid

Is it the same motherboard and power supply?

new motherboard same power supply corsair 700w

[Spritzup]

Older drive, would only be used as an interim solution anyways, but just wanted to know if I should be concerned.

Preclear_Results.txt

[Joe L.]

Older drive, would only be used as an interim solution anyways, but just wanted to know if I should be concerned.

I would not trust windows to write to those drives...  Apparently, it did not handle the writing very well.  It could have been anything if the disk is now plugged into different connectors, with better connections, etc it might just be fine.

 

Joe L.

[Spritzup]

I would not trust windows to write to those drives...  Apparently, it did not handle the writing very well.  It could have been anything if the disk is now plugged into different connectors, with better connections, etc it might just be fine.

 

Joe L.

 

Doesn't really instill me with confidence

[Joe L.]

I would not trust windows to write to those drives...  Apparently, it did not handle the writing very well.  It could have been anything if the disk is now plugged into different connectors, with better connections, etc it might just be fine.

 

Joe L.

 

Doesn't really instill me with confidence

I know... but you had two different disks with similar symptoms.  More than likely on different cables connected to different ports on your old MB.  The only common factors are

Your old MB.

Your old power supply

MS-Windows

 

Which one did you have the most confidence in?

 

Joe L.

[Spritzup]

I would not trust windows to write to those drives...  Apparently, it did not handle the writing very well.  It could have been anything if the disk is now plugged into different connectors, with better connections, etc it might just be fine.

 

Joe L.

 

Doesn't really instill me with confidence

I know... but you had two different disks with similar symptoms.  More than likely on different cables connected to different ports on your old MB.   The only common factors are

Your old MB.

Your old power supply

MS-Windows

 

Which one did you have the most confidence in?

 

Joe L.

 

Now I'm confused... my preclear results are for only one disk, and this is the first time I've posted here... I think you have me confused... or I'm confused.

[Joe L.]

I would not trust windows to write to those drives...  Apparently, it did not handle the writing very well.  It could have been anything if the disk is now plugged into different connectors, with better connections, etc it might just be fine.

 

Joe L.

 

Doesn't really instill me with confidence

I know... but you had two different disks with similar symptoms.  More than likely on different cables connected to different ports on your old MB.   The only common factors are

Your old MB.

Your old power supply

MS-Windows

 

Which one did you have the most confidence in?

 

Joe L.

 

Now I'm confused... my preclear results are for only one disk, and this is the first time I've posted here... I think you have me confused... or I'm confused.

My mistake.  I confused you with a prior user.

[Spritzup]

Perfect, thank you for clearing that up.  Any comment on my results though?

[Joe L.]

I would not trust windows to write to those drives...  Apparently, it did not handle the writing very well.  It could have been anything if the disk is now plugged into different connectors, with better connections, etc it might just be fine.

 

Joe L.

 

Doesn't really instill me with confidence

I know... but you had two different disks with similar symptoms.  More than likely on different cables connected to different ports on your old MB.   The only common factors are

Your old MB.

Your old power supply

MS-Windows

 

Which one did you have the most confidence in?

 

Joe L.

 

Now I'm confused... my preclear results are for only one disk, and this is the first time I've posted here... I think you have me confused... or I'm confused.

Looks like 1 sector re-allocated, 5 pending, but were written in place and not re-allocated, and you are using an OLD version of the pre-clear script.

Type

preclear_disk.sh -v

to see the version.

 

Other than that, the disk does not look too bad.  Spin up time getting worse, and has had to re-try 41 times out of 11,566, so perhaps bearings showing some wear.

[ricsouthcott]

just a update Joe L and prostuff1   i did a 2nd preclear on the 2 drives and both cleared prefect  so i think it was windows  when i was taking the movies off the drive i had a problem with a couple of the movies they would copy off , thanks for ya help

[Rajahal]

Hi Joe,

 

I have a total of 15 3TB Hitachi CoolSpin drives that I'm burning in for a client.  I figured I would take the opportunity to help test out preclear1.12beta.  I believe this post contains all the info you asked of other uses, but let me know if there's anything else you need.  Below are the results for 8 of the 15 drives.  These drives were divided into groups of four and precleared simultaneously using two different test servers.  The hardware of each server is described below.  Both servers were running unRAID 5.0-beta10.

 

 

Here's the time to completion for each drive:

 

Drive   Duration   Server Name

6H9A 48:49:35  Thailand

7MGA 47:41:13  Thailand

GY8A 48:28:24  Thailand

6AKA 48:55:51  Thailand

DYPA 41:29:50  Prototype

GE4A 42:06:13  Prototype

HSAA 40:28:49  Prototype

HXBA 43:02:11  Prototype

AVG:    44.6 hours

 

'Thailand' is a Supermicro X7SLA-H with a built-in Atom CPU and 2 GB of DDR2 533 RAM (2 x 1GB).  'Prototype' is a ZOTAC GF6100-E-E with a Sempron 140 CPU and 2 GB of DDR2 800 RAM (1 x 2GB).  Here's the averages across each set of hardware:

 

Thailand: 47.75 hours

Prototype: 41.5 hours

 

Interesting!  It seems that the slower CPU and RAM in 'Thailand' are enough of a difference to account for an extra 6.25 hours of preclear time!  I was wondering if 2 GB of RAM would be enough to simultaneously preclear four 3TB drives.  It appears it is enough, but slower RAM may contribute to longer preclear times.

 

I'll preclear some of the remaining drives on 'Prototype' with 4 GBs of DDR2 800 RAM (2 x 2GB) installed to see if there's any difference.  Unfortunately 'Thailand' is already at its maximum RAM capacity.

 

Edit: Oops, forgot to included the preclear reports.  They are now attached to this post.  I had to manually enter the .txt file extension for each one...didn't preclear used to take care of that by itself?  Also, the text files don't seem to align themselves nicely like they used to.  They now look like this:

 

 

Is that the fault of the preclear script, or my Win7 computer?

preclear_reports.zip

[madburg]

Drive   Duration   Server Name

6H9A 48:49:35  Thailand

7MGA 47:41:13  Thailand

GY8A 48:28:24  Thailand

6AKA 48:55:51  Thailand

DYPA 41:29:50  Prototype

GE4A 42:06:13  Prototype

HSAA 40:28:49  Prototype

HXBA 43:02:11  Prototype

AVG:    44.6 hours

 

Hey Raj, how many cycles for the times above?

[Octavius]

Hi all,

 

With the help of many of your community's posts, I've just build my first unRAID server. I was looking for something to replace my inadequate DNS-323, and I really liked everything that unRAID had to offer. Thank you all so much for being so helpful to the users and for answering their questions. It has made it that much easier for me to make the plunge (especially since while I'm quite the Linux noobie).

 

I've precleared to WDD HDD's twice, and both times on each drive it has said that the preclear was successful. I tried reading the output results but I couldn't really make heads or tails over much of what was in there.

 

Can someone please look through my log and tell me:

 

1) What specifically I should be looking for in determining if someone is wrong with an HDD

2) Whether or not my HDDs are good to go

 

Thanks in advance! Looking forward to getting more familiar with unRAID and in being a contributing member to the unRAID community!

 

I just attached the 2nd half of the tests, let me know if the first half is needed.

preclear_results_002.txt

[Rajahal]

Drive   Duration   Server Name

6H9A 48:49:35  Thailand

7MGA 47:41:13  Thailand

GY8A 48:28:24  Thailand

6AKA 48:55:51  Thailand

DYPA 41:29:50  Prototype

GE4A 42:06:13  Prototype

HSAA 40:28:49  Prototype

HXBA 43:02:11  Prototype

AVG:    44.6 hours

 

Hey Raj, how many cycles for the times above?

 

1 cycle each.  Running multiple cycles by using the -c flag would skew the data quite a bit, since on the 2nd, 3rd, etc. pass the previous cycle's post-read doubles as the current cycle's pre-read.  I will be running a second cycle on each drive (not using the -c flag) at a later date, so I'll report those numbers once I have them.  I'm currently setting up three of my test servers with 2-3 drives each to test out the SIL3132 controller as well as the Biostar A880G+ motherboard for 3TB drive compatibility.

[Rajahal]

Hi all,

 

With the help of many of your community's posts, I've just build my first unRAID server. I was looking for something to replace my inadequate DNS-323, and I really liked everything that unRAID had to offer. Thank you all so much for being so helpful to the users and for answering their questions. It has made it that much easier for me to make the plunge (especially since while I'm quite the Linux noobie).

 

I've precleared to WDD HDD's twice, and both times on each drive it has said that the preclear was successful. I tried reading the output results but I couldn't really make heads or tails over much of what was in there.

 

Can someone please look through my log and tell me:

 

1) What specifically I should be looking for in determining if someone is wrong with an HDD

2) Whether or not my HDDs are good to go

 

Thanks in advance! Looking forward to getting more familiar with unRAID and in being a contributing member to the unRAID community!

 

I just attached the 2nd half of the tests, let me know if the first half is needed.

 

Welcome, and congrats on building your first server!  If you are anything like the rest of us, it won't be your last

 

Interpretting SMART reports can be somewhat involved, but Joe L. (the author of the preclear script) has created this nifty output report that makes it much easier.  Each of your drives will have a report that looks like this:

 

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  No SMART attributes are FAILING_NOW

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  0 sectors were pending re-allocation before the start of the preclear.

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  0 sectors were pending re-allocation after pre-read in cycle 1 of 1.

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  0 sectors were pending re-allocation after zero of disk in cycle 1 of 1.

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  0 sectors are pending re-allocation at the end of the preclear,

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:    the number of sectors pending re-allocation did not change.

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  0 sectors had been re-allocated before the start of the preclear.

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  0 sectors are re-allocated at the end of the preclear,

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:    the number of sectors re-allocated did not change.

 

What this does is run a SMART report before the preclear cycle, then another one after the preclear cycle.  It then subtracts the before values from the after values and shows you the difference (the bold numbers above).  In this case, there were no differences, so the numbers are all zero.  This is a good thing.  Numbers greater than zero generally indicate some sort of mechanical problem with a drive, but they don't necessarily mean that the drive must be replaced immediately.  You have to evaluate the numbers on a case-by-case basis.  This is a topic that you'll learn about little bit at a time as you spend more time on these forums, don't try to digest it all at once.  There are a lot of parameters involved in assessing a hard drive's health.

 

In your case, all of your drives look perfectly healthy.

[madburg]

Hey Raj, how many cycles for the times above?

 

1 cycle each.  Running multiple cycles by using the -c flag would skew the data quite a bit, since on the 2nd, 3rd, etc. pass the previous cycle's post-read doubles as the current cycle's pre-read.  I will be running a second cycle on each drive (not using the -c flag) at a later date, so I'll report those numbers once I have them.  I'm currently setting up three of my test servers with 2-3 drives each to test out the SIL3132 controller as well as the Biostar A880G+ motherboard for 3TB drive compatibility.

 

Thanks,

 

I just finished preclearing (3) 3tb drives as well one being a Hitachi 7K and the other two Hitachi 5K drives. All at the same time. To give you some stats. (all precleared 2-Cycles successfully)

 

1-Cycle:

7K = 30:40:00

5K #1 = 37:36:30

5k #2 = 37:45:16

 

Xeon in a SuperMicro MB, all three drives on one 6Gbps LSI SAS Controller. Using Preclear 1.12beta

 

[SSD]

Edit: Oops, forgot to included the preclear reports.  They are now attached to this post.  I had to manually enter the .txt file extension for each one...didn't preclear used to take care of that by itself?  Also, the text files don't seem to align themselves nicely like they used to.  They now look like this:

 

 

Is that the fault of the preclear script, or my Win7 computer?

 

Try Notepad2 a lightweight text editor.  Does not even need to be installed.

[Octavius]

Hi all,

 

With the help of many of your community's posts, I've just build my first unRAID server. I was looking for something to replace my inadequate DNS-323, and I really liked everything that unRAID had to offer. Thank you all so much for being so helpful to the users and for answering their questions. It has made it that much easier for me to make the plunge (especially since while I'm quite the Linux noobie).

 

I've precleared to WDD HDD's twice, and both times on each drive it has said that the preclear was successful. I tried reading the output results but I couldn't really make heads or tails over much of what was in there.

 

Can someone please look through my log and tell me:

 

1) What specifically I should be looking for in determining if someone is wrong with an HDD

2) Whether or not my HDDs are good to go

 

Thanks in advance! Looking forward to getting more familiar with unRAID and in being a contributing member to the unRAID community!

 

I just attached the 2nd half of the tests, let me know if the first half is needed.

 

Welcome, and congrats on building your first server!  If you are anything like the rest of us, it won't be your last

 

Interpretting SMART reports can be somewhat involved, but Joe L. (the author of the preclear script) has created this nifty output report that makes it much easier.  Each of your drives will have a report that looks like this:

 

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  No SMART attributes are FAILING_NOW

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  0 sectors were pending re-allocation before the start of the preclear.

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  0 sectors were pending re-allocation after pre-read in cycle 1 of 1.

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  0 sectors were pending re-allocation after zero of disk in cycle 1 of 1.

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  0 sectors are pending re-allocation at the end of the preclear,

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:     the number of sectors pending re-allocation did not change.

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  0 sectors had been re-allocated before the start of the preclear.

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:  0 sectors are re-allocated at the end of the preclear,

Jul 26 13:46:10 Tower preclear_disk-diff[24306]:     the number of sectors re-allocated did not change.

 

What this does is run a SMART report before the preclear cycle, then another one after the preclear cycle.  It then subtracts the before values from the after values and shows you the difference (the bold numbers above).  In this case, there were no differences, so the numbers are all zero.  This is a good thing.  Numbers greater than zero generally indicate some sort of mechanical problem with a drive, but they don't necessarily mean that the drive must be replaced immediately.  You have to evaluate the numbers on a case-by-case basis.  This is a topic that you'll learn about little bit at a time as you spend more time on these forums, don't try to digest it all at once.  There are a lot of parameters involved in assessing a hard drive's health.

 

In your case, all of your drives look perfectly healthy.

 

Awesome. Thanks very much for the warm welcome and the interpretation of the smart report. I'll keep an eye out for any anomalies!

 

Thanks again!

[Rajahal]

3 TB drive preclear test results, round 2 (all drives are 3 TB Hitachi CoolSpins, all test servers running unRAID 5.0-beta10, all running preclear 1.12 beta):

 

 

This time around I tested the SIL3132 SATA controller.  I'm using the Monoprice controllers, but any 2 port PCIe x1 SIL3132 controller should have similar if not identical results.

 

Drive   Duration   Server

6E1A 54:47:52   Thailand

B1NA 53:24:05   Thailand

GE5A 46:49:55   Prototype

GEUA 47:30:24   Prototype

AVG:   50 hours

 

'Thailand' is a Supermicro X7SLA-H with a built-in Atom CPU and 2 GB of DDR2 533 RAM (2 x 1GB).  'Prototype' is a ZOTAC GF6100-E-E with a Sempron 140 CPU and 2 GB of DDR2 800 RAM (1 x 2GB).  Both servers were configured with a single SIL3132 card plugged into a PCIe x16 port.  My first round of testing resulting in an average preclear time of 44.6 hours when using the motherboards' onboard SATA ports.  The SIL3132 controllers appear to slow things down by over 6 hours, with an average of 50 hours per cycle of preclear.

 

Average per server:

 

Thailand: 53.5 hours

Prototype: 46.5 hours

 

Yet again, the slower core components appear to result in significantly slower preclear times, a difference of 7 hours.

 

I think these results make it pretty clear that if you are looking for the fastest preclear times SIL3132 controllers should be avoided.  However, if you aren't concerned with preclear taking a couple extra hours to complete, there's nothing wrong with using these controllers.  Seeing as they are very inexpensive, they are a great way to eek out a few more SATA ports from an otherwise full motherboard.  They are compatible with PCIe x1, x4, x8, and x16 slots, so they are very versatile.  There are also PCI versions of the card, but I would expect these to have even worse performance due to PCI bus limitations.

 

---

 

I also tested a new motherboard for 3 TB drive compatibility: the Biostar A880G+.  Here are those results:

 

 

Drive   Duration   Server

HXBA   41:58:44   Dev

EL8A   42:25:21   Dev

G83A   41:36:09   Dev

HZMA   42:44:09   Dev

AVG:     41.5 hours

 

'Dev' is a Biostar A880G+ with a Sempron 140 CPU and 2 GB of DDR3 1333 RAM (1 x 2GB).  Average preclear speed is exactly the same as the 'Prototype' test server, which has very similar specs.  I may run a further test of the SIL3132 card attached to this board, but I'm not sure it is really necessary, since the results should be the same as the 'Prototype' results above.

 

---

 

Preclear reports for all of the above are attached below.  By the way, I took bjp999's advice and tried a different text editor.  I already had notepad++ installed, so I used that and the preclear reports were formatted correctly.  Not sure why Window's built in notepad no longer displays these correctly, but I suppose it is no big deal.

 

So far I've determined that all of the following hardware are compatible with 3 TB drives and preclear 1.12 beta:

Motherboards

Supermicro X7SLA-H

ZOTAC GF6100-E-E

Biostar A880G+

 

SATA Controller

SIL3132

 

The remaining items I plan to test over the next week are:

 

Motherboards

Supermicro X8SIL-F-O

ECS A785GM-M7 (candidate for new 15 drive budget board)

Biostar A760G M2+

 

SATA Controller

Supermicro AOC-SASLP-MV8

 

I will continue to post my results for the above hardware in this thread.  Once I've finished with all of this testing I will compile all of the data and results into a single post on the GreenLeaf hardware blog.

preclear_reports_round_2.zip

[Traxxus]

So I have issues with a new EARS drive I'm adding to the server.  No jumper, upgraded to 4.7 recently in order to take advantage of the new AF formatting, and I do have it enabled in settings.  

 

Syslog is 540 MB in size, mostly the same error repeating over and over with this

 

Aug 1 05:11:04 Tower kernel: sd 3:0:0:0: [sdc] Unhandled error code
Aug 1 05:11:04 Tower kernel: sd 3:0:0:0: [sdc] Result: hostbyte=0x04 driverbyte=0x00
Aug 1 05:11:04 Tower kernel: sd 3:0:0:0: [sdc] CDB: cdb[0]=0x2a: 2a 00 e2 25 cf 68 00 04 00 00
Aug 1 05:11:04 Tower kernel: end_request: I/O error, dev sdc, sector 3794128744

 

I took the end and beginning of the syslog and attached it.  

The preread appeared to go fine, took 8 hours, then failed on the write portion I presume.  Smart test returns this.  (full text)

smartctl 5.39.1 2010-01-28 r3054 [i486-slackware-linux-gnu] (local build)
Copyright (C) 2002-10 by Bruce Allen, http://smartmontools.sourceforge.net

Smartctl: Device Read Identity Failed (not an ATA/ATAPI device)

A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options.

 

I double checked the cables and they look good, assuming that isnt the problem since the read portion appeared to go fine.  

 

Oh, and the preclear script returned this during the test.  

 

Sorry disk sdc MBR could not be precleared
== out4= 00000
== out5= 00000
============================================================================
0+0 records in
0+0 records out
0 bytes (0 B) copied, 1.7715e-05 s, 0.0 kB/s
0000000

== out4= 00000
== out5= 00000
============================================================================
0+0 records in
0+0 records out
0 bytes (0 B) copied, 1.7715e-05 s, 0.0 kB/s
0000000

 

Elapsed Time:  7:40:08
dd: writing `/dev/sdc': Input/output error
1+0 records in
0+0 records out
0 bytes (0 B) copied, 0.00104791 s, 0.0 kB/s

 

Edit:  Think I fucked up that code somewhere, putty was being funny about c/p some of it. 

syslog.txt

smart.txt

[Joe L.]

Looks like the disk stopped responding to all commands from the OS.  

The "dd" commands stating zero records read are a big clue.

 

In past cases, it takes a power cycle to get it to respond again. (power down server, then restore power, then try smartctl report again)

 

What version of the preclear script are you using?  Early versions attempt to read past the end of the disk and some disks react by locking up. (firmware on the disk is probably buggy in not being able to handle this)

 

To see the version of the preclear script, type:

preclear_disk.sh -v