But Joe do you really think this is a power issue? I was able to copy a bunch of files off of disk 6.

I really do not know.  You have a lot of un-related problems with a lot of your disks...  One common cause would be noisy power (un-regulated since it is past its limits)

 

You can decide.  For me you are way past the limits of that supply.  Perhaps the task of moving the disk heads to the innermost tracks on ALL the disks takes more power than reading the outermost tracks.  That might explain your failure to complete the re-construction.  As you said, a  temporary fix is easy.  Move some of the disks to an alternate supply.  then you can test if it makes a difference. 

 

Get RID of the 40 conductor IDE cables.  Get good quality 80 conductor cables.  (NOT ROUND cables)  All these are critical to reading the disks reliably.

 

Joe L. 

Ok moving 5 drives to second power supply:

 

Cooler Master Extreme Power plus 460w says it has 2 +12V rails each capable of 18A

 

5 drives at 3A = 15A on Cooler Master

6 drives at 3A = 18A on Enermax

 

That should be ok.

 

Lets see if this works.

Cooler Master Extreme Power plus 460w says it has 2 +12V rails each capable of 18A

 

5 drives at 3A = 15A on Cooler Master

6 drives at 3A = 18A on Enermax

 

That should be ok.

 

Lets see if this works.

 

You don't have much headroom at this Enermax because 12V1 usually is used by MB including those SATA/Molex connectors. This PSU is designed for gaming SLI that is why it has two PCI-e power connector and if you look at vendor's web site that is what they try to highlight. Nowadays high-end powerful video cards need extra power input. In the mean time, depends on how long you own this PSU and how you use it before, those components in a PSU will age and PSU will not be as efficient as before (otherwise all PSU vendors will be out of business soon), what spec say is for brand new unit.

 

There are many PSU discussions before maybe you should look it up.

 

Cooler Master Extreme Power plus 460w says it has 2 +12V rails each capable of 18A

 

5 drives at 3A = 15A on Cooler Master

6 drives at 3A = 18A on Enermax

 

That should be ok.

 

Lets see if this works.

 

You don't have much headroom at this Enermax because 12V1 usually is used by MB including those SATA/Molex connectors. This PSU is designed for gaming SLI that is why it has two PCI-e power connector and if you look at vendor's web site that is what they try to highlight. Nowadays high-end powerful video cards need extra power input. In the mean time, depends on how long you own this PSU and how you use it before, those components in a PSU will age and PSU will not be as efficient as before (otherwise all PSU vendors will be out of business soon), what spec say is for brand new unit.

 

There are many PSU discussions before maybe you should look it up.

 

He had all 11 on the Enermax... this is an improvement... and a test.

He had all 11 on the Enermax... this is an improvement... and a test.

 

Hmmm.... I wonder if those disks have high "spin retry count" from SMART.

 

 

disk5 WDC_WD20EARS-00M_WD-WCAZA0985379 22°C 1,953,514,552 1,933,671,552 33 2,078,330 0

disk6 ST3500841A_3PM079GN 0°C 488,386,552 245,981,228 1,509,788 7 27,954,855

 

Total size: 1,953,514,552 KB

Current position: 624,988,168 (31.9%)

Estimated speed: 30,790 KB/sec

Estimated finish: 719.1 minutes

 

Folks I do not think this is a power or data cable issue any more. The rebuild is now showing the same amount of errors on drive 6. I think we have bad timing with drive 5 and drive 6 going out at almost the same time. I replaced another 500gig some time ago with 2TB with the parity drive also being swapped out - all these drives were bought at about the same time - it figures that they all fail about the same time - the drives are at least 3 years old.

 

Thanks for all the help guys - I will replace the power supply in any case just to be safe. If you have a specific model that is a good fit/price for this use I would appreciate the recommendation. Is this a good one? US Power and Cooling - Silencer 750 Quad Black

 

I will yank out the remaining IDE PATA drives - who needs this hassle with 80wire cables etc. I only have 4x200gig PATA drives left in any case - one more 2TB drive will take care of that.

 

I will grab what ever else I can from drive 6 and 5 and then forget about them.

 

So I have the following questions:

 

1. How can I mount and read a reiserfs on Windows 7?

- Looking at using this: http://p-nand-q.com/download/rfstool.html with http://yareg.akucom.de/

2. What is the best way to get rid of this rebuild situation? And just create two fresh drive 5 and drive 6?

3. What is the best way to replace drive 7,8,9.10 (200gig) with one drive 2TB?

4. How often should I be running reiserfschk on my drives?

5. Is there a thread/FAQ that explains how to read the log files?

 

 

 

Thanks for all the help guys - I will replace the power supply in any case just to be safe. If you have a specific model that is a good fit/price for this use I would appreciate the recommendation. Is this a good one? US Power and Cooling - Silencer 750 Quad Black

Yes, it has a single 60 Amp 12 volt rail.  It would work perfectly.

I will yank out the remaining IDE PATA drives - who needs this hassle with 80wire cables etc. I only have 4x200gig PATA drives left in any case - one more 2TB drive will take care of that.

Your choice.  From a power perspective a single 2TB drive will draw 1/4 of the power as four 200 Gig drives, and likely less, as older drives were not as "green".

 

Then you can have all the fun with poor quality SATA connections and loose connectors. (nothing is perfect)  I have one unRAID server with 10 IDE drives, on flat 80 conductor cables, and no problems.  The disks are connected to a pair of 350 Watt power supplies.  True they are not as efficient as today's drives, but they have been in operation for over 5 years.

Joe any chance of some more answers to my many questions ?

 

Thanks

 

Joe any chance of some more answers to my many questions ?

 

Thanks

I know there are read-only drivers to read reiserfs on windows.  I have absolutely no idea if they work on windows-7.

 

To get rid of the additional drives stop the array, un-assign them, and with the array stopped type

initconfig

on the linux command line.

Answer "Yes" to its prompt.  (Capital "Y", lower case "es")

If you did it correctly, when you refresh the main page on the web-interface, all the drives should show as "BLUE" indicators.

 

You should never need to run reiserfsck on your drives.

 

There is no "thread" on how to read log files.  Google is your friend.  Many lines are simply status messages.  Even "errors" might be completely normal in the start-up of the server as the OS is probing for possible hardware and an error might just indicate you do not have specific hardware.

 

Joe L.

So here's what I am planning to do.

 

1. Leave the brand new disk5 - 2TB SATA in place

2. Take out the bad disk 6 PATA - add 2 TB SATA to replace it

3. Go to command line and type INITCONFIG -> type Yes

4. Do a Parity resynch

 

Then once everything is ok.

 

1. Use Midnight Commander to copy stuff off of 4x 200gig PATA drives to 2TB SATA drive in disk 6

2. Take out disks 7-10

3. Repepat step 3  - INITCONFIG.

 

End state

 

Server is pure SATA

Reduce number of drives from 11 to 6

 

 

Please correct/confirm

 

I will then:

Install new PC power and cooling power supply

and new drive cage I have on order

 

Joe, what line in sys or msg log shows that I am not using 80 wire PATA cable? Because when I look at the cables they all look like 80 wire cables.

Joe, what line in sys or msg log shows that I am not using 80 wire PATA cable? Because when I look at the cables they all look like 80 wire cables.

 

The cables you want are the 80pin FLAT variety.  They are better quality 99% of the time and are the ones recommended for use in an unRAID server.  The round version can give all sorts of problems.  I have an older test server that has a couple IDE drives in it and I find that hiding the flat cables is much easier and looks cleaner.  I never really understood the round IDE cable fad, as to me they just stuck out more and were harder to hide... I can bend and hide a flat cable almost anywhere I like.

So here's what I am planning to do.

 

1. Leave the brand new disk5 - 2TB SATA in place

2. Take out the bad disk 6 PATA - add 2 TB SATA to replace it

3. Go to command line and type INITCONFIG -> type Yes

4. Do a Parity resynch

 

Then once everything is ok.

 

1. Use Midnight Commander to copy stuff off of 4x 200gig PATA drives to 2TB SATA drive in disk 6

2. Take out disks 7-10

3. Repepat step 3  - INITCONFIG.

 

End state

 

Server is pure SATA

Reduce number of drives from 11 to 6

 

 

Please correct/confirm

 

I will then:

Install new PC power and cooling power supply

and new drive cage I have on order

 

Joe, what line in sys or msg log shows that I am not using 80 wire PATA cable? Because when I look at the cables they all look like 80 wire cables.

The command syntax in linux is case sensitive.  The command is

initconfig

not INITCONFIG

you probably know that, but others reading this thread in the future might not.

 

I'd leave the temporary supply you connected in place for now.  You may be way overloading the existing supply, even with 6 drives.

 

Joe - All else good? The last thing I want to happen is loose any of the other drives - those actually hold stuff I need.

 

BTW just for kicks I allowed the server to finish the current rebuild - 99.9% - and stuck!

 

I am tired of working on it - I am going to let it sit at 99.9% for a day or so to see what happens.

 

Yes - I will keep the second power on till new power supp;ly arrives.

Just for a laugh, take a look at my OCD schedule for moving a 2TB drive from my Buffalo NAS and adding that and another (new) 2TB drive to my unRAID server, which I started yesterday.

The drive currently in the Buffalo NAS needs to be the parity drive, so that complicates things:

 

Now: TOWER is 17 drives

 

Sunday:

-------

* Check power and SATA for new drive

* Parity check T2 AND TOWER

* Add monitor

 

Monday

------

* JUMPER PINS 7-8

* WD 2TB in to TOWER

* JUMPER PINS 7-8

* Preclear WD2TBEARS

* Add as cache drive

 

---------------------------------

Need to free up 465GB to proceed.

95GB free on share

380GB in Radio_files

= 465GB

---------------------------------

 

* Copy 1TB of movies from 1TB disk3 or disk6 to WD2TBEARS

* copy 320GB from LSLIVE to WD2TBEARS

* Make list of unprotected files on WD2TBEARS

* Put 1TB drive in LSLIVE

* Copy data from LSPRO to LSLIVE (913-465GB = 465GB) non-Movies data to LSLIVE

* StevieWonder LSLIVE

* Set up LSLIVE like LSPRO

Remaining: 1TB movies on LSPRO

Space on WD2TBEARS: 640GB

* Copy 640GB Movies from LSPRO to WD2TBEARS

* Copy 320GB Movies from LSPRO to Spare: 320GB SATA from LSLIVE (NTFS?)

* Back up important data from LSPRO to T2 and FW11S. Test it. Test files on LSLIVE

* CHECK Radio_files ON T2 - these will go from LSPRO

Now 2TB Seagate is empty

* Put 2TB Seagate in TOWER -- replace existing parity drive

Now 1.5TB free on TOWER old 7200 SG 1.5TB drive and 500GB free on LSLIVE 1TB

* Copy 1.5TB movies from WD2TBEARS to 7200 SG 1.5TB drive and 500GB to LSLIVE 1TB

 

* Allocate WD2TBEARS from cache to array drive

* Copy 500GB movies from LSLIVE 1TB to WD2TBEARS

Now 1.5TB free on WD2TBEARS

* Copy Radio_files files from T2 to LSLIVE

 

* Move data from 17th TOWER drive to WD2TBEARS

 

* Move 17th drive to T2

 

* Move (up to) 1TB movies from T2 to WD2TBEARS

* Move (up to) 1.5TB TV Blu-ray from TOWER to T2 (all BSG?)

Joe what about this:

 

Remove one or more data disks

In this case the missing disk(s) will be identified. If there is only one missing disk when you start the array it will be marked as failed. All data disks will be exported (including the missing one), but the system will be running unprotected; that is, if a disk fails you will lose data.

If there are two or more missing disks, you can not start the array. In this case you must either put the disks back, or click Restore on the Main page to reset the configuration.

 

Is what you said to do the same as this? ie: end result? is one method preferred over the other?

 

 

Result from this command attached:

smartctl -a -d ata /dev/hdc > /boot/config/smarterror.txt

 

Does it show a bad drive?

smarterror.txt

 

 

Result from this command attached:

smartctl -a -d ata /dev/hdc > /boot/config/smarterror.txt

 

Does it show a bad drive?

Yes.  The drive is bad

 

416 sectors have been re-allocated, and 4780 are pending re-allocation.   (typically there are only a few thousand in the spare pool, probably less in a smaller drive like this)

 

RMA the drive.... (if possible)  I see from the power on hours the drive is over 3 years old.

 

 5 Reallocated_Sector_Ct   0x0033   090   090   036    Pre-fail  Always       -       416

197 Current_Pending_Sector  0x0012   001   001   000    Old_age   Always       -       4780

Joe this method or your method? or are they the same?

 

Joe what about this:

 

Remove one or more data disks

In this case the missing disk(s) will be identified. If there is only one missing disk when you start the array it will be marked as failed. All data disks will be exported (including the missing one), but the system will be running unprotected; that is, if a disk fails you will lose data.

If there are two or more missing disks, you can not start the array. In this case you must either put the disks back, or click Restore on the Main page to reset the configuration.

 

Is what you said to do the same as this? ie: end result? is one method preferred over the other?

took a couple of days off from this project to just get away from it.

 

followed steps.

 

Array is up.

 

Parity sync is ongoing as of 4:35pm EST 11/4/10.

 

Thanks all.

Parity completed - all seems fine.

 

Thanks again to everyone.

One final post

- looks like disk 5 did get rebuilt - I do not know how valid the data is its a bunch of DVDs and I have not watched every one of them right through so I do not know if some of them are corrupted. But the ones I checked seem to be fine.