Jump to content
jimwhite

Anyone here bit by the "Seagate Barracuda 7200.11" bug?


49 posts in this topic Last Reply

Recommended Posts

I've been following this thread: http://www.msfn.org/board/index.php?showtopic=128092 for a while now, and wondering if and when this bug will bite me as I have 12 of the 1tb's in my unRAID box and 8 of the 500mb guys in my desktops.... Since there are probably a LOT of Seagate drive owners who are unRAID users, or vice versa, I wanted your input.

 

Share this post


Link to post

Hmmmm..... looks like NOBODY has EVER lost a seagate drive???  ???

 

I seriously don't think a whopping sampling size of 6 is sufficient basis for that. ::)

Share this post


Link to post

We haven't seen any problems with the 1TB drives, probably have tested around 100 of them.  But the 1.5TB hard drives definitely have problems - in fact we took them down from our Hard Drives page at this time:

http://lime-technology.com/?page_id=37

 

What we see are I/O time-outs and occasionally complete drop outs.  Interesting, firmware release "CC1J" does seem to work, but we recently got a batch of "CC1H" which is terrible.  Seagate is no help and claim that drives with CC1H can not be upgraded to CC1J.  So at this time we can not recommend them, which is unfortunate because I know many people using unRAID are also using this drive.  It's also possible that we just got a "bad batch" and it has nothing to do with the firmware, but that would be a first.

Share this post


Link to post

I have 2 Seagate 7200.11 1T drives.  They are working fine.  One is my parity disk.  The only problem is that the smart attributes indicate between 10-15 spinup retries.  I've observed this getting incremented every week or two.  I have seen other users smart output as well, and observed these drives often have non-zero spin retry attributes.  Not sure if this will be problematic in the future. 

 

I have a bunch of 7200.10 drives (750G and 500G), and NONE of them have this issue.  Seagate used to enjoy a sterling reputation, but I think it is getting tarnished.  Unrelated to this, I have switched to the WD drives primarily due to the lower power consumption, but also I believe that the fact that they spin slower will have a positive impact on longevity.  The fact that faster drives are not needed for a media application like unRAID also influenced the decision.  I am very pleased with them.

Share this post


Link to post

...The only problem is that the smart attributes indicate between 10-15 spinup retries.  I've observed this getting incremented every week or two. 

 

Seagate used to enjoy a sterling reputation, but I think it is getting tarnished.   

 

I'm experiencing the same thing.   No failures yet, but I'm hawking those drives.

Share this post


Link to post

I purchased a total of 8 of these 1.5TB.  After issues ranging from ever increasing SMART errors, sector reallocations, firmware issues and not the greatest support from Seagate, I've pulled the all from my unRAID servers.  I just don't trust them... I will wait until WD releases a green power 1.5TB.

 

Regards,  Peter

Share this post


Link to post

Excluding my cache drive all four of my drives are Seagate 7200.11:

One 1TB ST31000350AS, firmware SD15, made in Thailand.  This was refurb drive from RMA that replaced a drive that died in another machine (without going into details, I'll assure you that the failure was not Seagate's issue).

Three 1.5TB ST31500350AS, firmware SD1A.  Two were flashed up from SD17, one from SD18, before deployment.

 

All of the drives come up relatively clean on SMART status, but all have some number of "high fly writes".  A couple of the 1.5TB drives seem to see these high fly writes increasing slowly but surely over time.  My third 1.5TB drive also repeatedly suffers from emask frozen errors (at least one or two per day, usually around spinups / spindowns but one time when the array was being stressed) requiring a "hard resetting link" but the cause of this is unclear to me at this time and so far has not led me to believe it is on the verge of failure.  FWIW, this happens to be the drive most recently purchased and flashed up from SD18.

 

In general, if I didn't know better and aside from the ritual of flashing them before installation, the drives seems perfectly fine to me and perform very well.  They do have active cooling and run around 39-45C with ambient 23C-26C (74F-80F).  At the same time, I'm not rushing to expand capacity until absolutely needed since I want to see how all this will shakeout...

 

edit:

I will wait until WD releases a green power 1.5TB.

 

Or maybe a 2TB, if you believe the rumors;)

Share this post


Link to post

Or maybe a 2GB, if you believe the rumors;)

 

I guess we will know later this week.  Wow!

 

Regards,  Peter

Share this post


Link to post

I've been following this thread: http://www.msfn.org/board/index.php?showtopic=128092 for a while now, and wondering if and when this bug will bite me as I have 12 of the 1tb's in my unRAID box and 8 of the 500mb guys in my desktops.... Since there are probably a LOT of Seagate drive owners who are unRAID users, or vice versa, I wanted your input.

 

 

I think I have just had a drive die due to this, it was a 1TB unit with SD15 firmware made in China, I think I bought it in July'08.  See my post:

 

http://lime-technology.com/forum/index.php?topic=3142.0

 

Stephen

 

Share this post


Link to post

So when does everyone bet the 1tb Bargains prices will kick in... try to compensate for the bad press...  Like to get so,e $79 1TB drives... Would be willing to do the BIOS update if they needed em...

 

Seagate just lost the CEO and President and laid off a bunch of people this past week (40% of internal IT US Staff if rumors are true) <- Unrelated to Issue

 

Rough week or two all the way around for Seagate...

 

SW2

Share this post


Link to post

If your drive is bricked, and you want to recover the data, here is a working procedure:

 

The Solution for 7200.11 Seagate HDDs:

 

Attention: This guide isn't intent for unexperienced users! Is very easy to KILL your HD if you don't follow the instructions with extreme caution. You have been warned! I'm NOT responsible for any damage or data loss it can causes.

 

You'll need:

 

- 1 (one) RS232 to TTL (I recommend those with +5V so you can use power from power supply PC).

- 1 (one) RS232 cable

- 1 (one) PC with Terminal software (like HyperTerminal) and RS232 port

- 4 (four) wires (for TX, RX, GND and +5V)

- Some jumpers, or similar (to link TX and RX on Seagate)

- Soldering Iron (optional)

- A fixed firmware from Seagate (to be released around 01/20)

- A torx tool

 

TX and RX connections from a SATA Seagate 3.5":

 

 

Connect TX and RX ports with RS232 to TLL adapter, you'll need something like jumpers in order to do that, or soldering directly to the PCB board. Connect GND and +5V to PC power supply, or similar.

 

Keep in mind RED wire from supply is +5V, and green or black are GNDs (yellow is +12V so don't use it).

 

Run HyperTerminal (or your favourite terminal software).

 

Config the RS232 port to: 38400 8 N 1

 

 

Fixing 0 LBA error:

 

To correct 0 LBA error, you'll need to do a quick format in order to regenerate the partition, just with m0,2,2 and the rest can be left blank by putting only commas or filling out with zeros. At the end you need to use a "ValidKey" code, or 22. So we'll have to enter:

 

 

F3 T>m0,2,2,0,0,0,0,22 (enter)

 

 

Now wait 15~30 seconds until something like this appears (it will vary a bit):

 

 

 

Max Wr Retries = 00, Max Rd Retries = 00, Max ECC T-Level = 00, Max Certify Rewrite Retries = 0000

 

User Partition Format 5% complete, Zone 00, Pass 00, LBA 00008DED, ErrCode 00000080, Elapsed Time 0 mins 05 secs

 

User Partition Format Successful - Elapsed Time 0 mins 05 secs

 

 

ONLY after the message like above, enter:

 

 

F3 T>/2 (enter)

 

F3 2>Z (enter)

 

Spin Down Complete

Elapsed Time 0.147 msecs

 

 

NOW you can turn off everything.

 

Conglatulations, you done it! (Goodbye money! Oh my! I could be rich!) 

 

 

 

Fixing BSY or CC errors aka BUSY (this will be kinda tough! I recommend a lot practice before to trying):

 

First you need to remove the PCB from HDA case, after that, connect RX & TX cables, verify that all is correct, and then power supply to PCB.

 

Explanation: To drives that hangs all the time in "busy" and on terminal (for example: LED: 000000CC FAddr: 0024A051 error codes), in which it blocks the terminal and it doesn't respond to any command, first we'll have to block off the time reading from plates on HD, in other words, power off your HD and DISCONNECT the PCB from HDA, with PCB alone, power up and press CTRL+Z (to enter on CMD mode), now we'll need to access the Level 2:

 

 

F3 T>/2 (enter)

 

F3 2>

 

 

And enter (to stop the spindle motor):

 

 

F3 2>Z (enter)

 

Spin Down Complete

Elapsed Time 0.147 msecs (the time will vary here)

F3 2>

 

 

 

Now the hard part. Pratice a lot without any power (of course), before even trying anything here.

 

Then mount the PCB back to HDA case, and DO NOT power off or disconnect the PCB from RX/TX (RS232) and power supply !

 

Now enter (to start the motor):

 

 

F3 2>U (enter)

 

 

You should get something like this:

Spin Up Complete

Elapsed Time 7.093 secs

F3 2>

 

Then go to Level 1 (by typing /1):

 

 

F3 T>/1 (enter)

 

 

Now lets do a S.M.A.R.T. erase (create S.M.A.R.T. sector):

 

 

F3 1>N1 (enter)

 

 

Now turn off (or power off) the drive. At this point I don't know yet if the drive power off alone after N1 command or not (I'm still need to order my RS232 to TTL adaptor). So, I'll just mention that as "Power OFF/ON" process.

 

G-List Erase (cert the reserved cyl):

 

 

F3 T>i4,1,22 (enter)

 

 

Power OFF/ON

 

Partition regeneration:

 

 

F3 T>m0,2,2,,,,,22 (enter)

 

 

You should get something like (in around 15~30 seconds):

 

 

Max Wr Retries = 00, Max Rd Retries = 00, Max ECC T-Level = 14, Max Certify Rewrite Retries = 00C8

User Partition Format 10% complete, Zone 00, Pass 00, LBA 00004339, ErrCode 00000080, Elapsed Time 0 mins 05 secs

User Partition Format Successful - Elapsed Time 0 mins 05 secs

 

 

If you didn't get the msg above, then you skipped something important, and turning off your drive now will render an unreadable drive, so be careful.

 

ONLY after the message like above, enter:

 

 

F3 T>/2 (enter)

 

F3 2>Z (enter)

 

Spin Down Complete

Elapsed Time 0.147 msecs

 

 

NOW you can turn off everything.

 

Conglatulations, you done it! (Goodbye money! Oh my! I could be rich!) 

 

Reference files (.fbr video [FlashBack] and pdf).

 

Credits goes to the peoples from:

http://yura.projektas.lt/files/seagate/720011_ES2.html

http://forum.hddguru.com/tutorial-resolve-...ies-t11031.html

http://www.elektroda.pl/rtvforum/viewtopic.php?p=6012404

http://www.elektroda.pl/rtvforum/topic1198189.html

 

Keep in mind I did my best to compile all this information and provide more and better details.

 

Gradius

 

Edit: the default speed is indeed 38400 bps

 

Share this post


Link to post

Has anybody gotten the firmware updates to work? I've been reading that a lot of people are having problems with it.

 

Edit: Seems like their knowledge base is down at the moment. There also also plenty of horror stories about Seagate customer service on their forums, not good for their reputation.

 

Edit2: From reading it seems that some of the firmware they release had problems, and killed some people's drives. Be careful when upgrading.

Share this post


Link to post

I flashed my 1.5tb a couple of weeks back and seems to be humming just fine.  It had the old firmware.

 

Im using mine as a parity drive.

Share this post


Link to post

Seagate has now acknowledged the problem and plan on releasing a firmware update for the affected disks next week. A serial number checker should be up on monday.

 

http://seagate.custkb.com/seagate/crm/selfservice/search.jsp?DocId=207931

 

sea321.png

 

I put together a short shell script to check each of my drives against the list and print out the model/serial/firmware of those matching the potentially affected SEAGATE drive models.

 

The output looks like this:

root@Tower:/boot# seagate_check.sh

[pre]The following SEAGATE disks may be in need of disk firmware upgrade.

    Device                    Model                  Serial    Firmware

  /dev/sdc              ST31500341AS                9VS069T8    CC1J

  /dev/sdd              ST31500341AS                9VS0G9D8    CC1J

  /dev/sdg              ST31000340AS                9QJ0JPJS    SX15

[/pre]

 

The script uses "smartctl" to grab the model, serial, and firmware from all the drives on your server.    If you are running version 4.4x or 4.5x you will probably need to download and install a missing library before "smartctl" will function.  See this post for details: http://lime-technology.com/forum/index.php?topic=2817.msg23548#msg23548

 

The script is here (and a zipped copy is attached)

#!/bin/bash
models="ST31000340AS
ST3750330AS
ST3640330AS
ST3500320AS
ST31500341AS
ST31000333AS
ST3640323AS
ST3640623AS
ST3320613AS
ST3320813AS
ST3160813AS
ST31000340NS
ST3750330NS
ST3500320NS
ST3250310NS
STM31000340AS
STM3750330AS
STM3500320AS
STM31000334AS
STM3320614AS
STM3160813AS"

for d in /dev/[hs]??
do
   diskdata=`smartctl -d ata -i $d 2>&1`
   echo "$diskdata" | grep "missing "
   if [ $? = 0 ]
   then
     echo "You will need to install the missing c++ library to run smartctl on your drives"
     exit 2
   fi
   disk_model=`echo "$diskdata" |  awk ' /Model:/ { print $3 }'`
   disk_serial=`echo "$diskdata" | awk ' /Serial/ { print $3 }'`
   disk_firmware=`echo "$diskdata" | awk ' /Firmware/ { print $3 }'`
   for m in $models
   do
      if [ "$disk_model" = "$m" ]
      then
         if [ "$title" = "" ]
         then
           printf "The following SEAGATE disks may be in need of disk firmware upgrade.\n"
           printf "%10s %20s    %20s    %s\n" "Device" "Model" "Serial" "Firmware"
           title="y"
         fi
         printf "%10s   %20s    %20s    %s\n" $d $disk_model $disk_serial $disk_firmware
         break;
      fi
   done
done
if [ "$title" = "" ]
then
  printf "You do not have any SEAGATE disks in the affected list needing firmware update"
fi

 

This may help some of you, especially if you have a lot of drives to check and need an easy way to collect the model/serial/firmware versions to use on the SEAGATE site to determine if your drives need firmware upgrade.

 

Use the model specific links on this page on the SEAGATE website to determine if your drives are affected.

 

I was lucky, none of my drives seem to have the buggy firmware.

 

Joe L.

Share this post


Link to post

Thanks for the info Joe. I used your info & found out 2 of my seagates need new firmware. I also found out that one of them had 7 spin retry counts, the other had 3 spin retry counts & one had 1 reallocated sectors.

 

I have been having problems of my box freezing up again when my drives go to sleep. Would the spin retry's have anything to do with that?

 

I have since told those drives to not go to sleep & see if that helps. I am now running 4.4.2

 

Thanks

Share this post


Link to post

So is there a way to flash these drives from the unRAID system, or do we have to pull, plug into Windows or something to flash with the Seagate update?

 

SW2

Share this post


Link to post

So is there a way to flash these drives from the unRAID system, or do we have to pull, plug into Windows or something to flash with the Seagate update?

 

SW2

 

Also is there a way to check firmware without pulling the HD's out. I see the script above, but that seems a little too complicated for me. Maybe some type of seagate tool?

 

Edit: Looks like SeaTools does this, I will try it out when I get home.

Share this post


Link to post

Anyone else notice they took down the fixed firmware at 8PM CST last night pending validaton?

 

Cheers,

Matt

 

 

Share this post


Link to post

So is there a way to flash these drives from the unRAID system, or do we have to pull, plug into Windows or something to flash with the Seagate update?

 

SW2

 

Also is there a way to check firmware without pulling the HD's out. I see the script above, but that seems a little too complicated for me. Maybe some type of seagate tool?

 

Edit: Looks like SeaTools does this, I will try it out when I get home.

If unzipping the attached file and invoking it by typing

seagate_check.sh

on the command line is a little too complicated, running seatools on your linux server, or flashing firmware on a hard drive is really going to tax your skills, regardless of the command involved.

 

Good luck...

 

Pull the drives out, read the label on the top of the drive.

 

Joe L.

Share this post


Link to post

So is there a way to flash these drives from the unRAID system, or do we have to pull, plug into Windows or something to flash with the Seagate update?

 

SW2

 

Also is there a way to check firmware without pulling the HD's out. I see the script above, but that seems a little too complicated for me. Maybe some type of seagate tool?

 

Edit: Looks like SeaTools does this, I will try it out when I get home.

If unzipping the attached file and invoking it by typing

seagate_check.sh

on the command line is a little too complicated, running seatools on your linux server, or flashing firmware on a hard drive is really going to tax your skills, regardless of the command involved.

 

Good luck...

 

Pull the drives out, read the label on the top of the drive.

 

Joe L.

 

Doh!

 

I thought it would have been more complicated then that!

 

Should I place the script in the root of my flash drive?

 

Thanks Joe!

Share this post


Link to post

So is there a way to flash these drives from the unRAID system, or do we have to pull, plug into Windows or something to flash with the Seagate update?

 

SW2

 

Also is there a way to check firmware without pulling the HD's out. I see the script above, but that seems a little too complicated for me. Maybe some type of seagate tool?

 

Edit: Looks like SeaTools does this, I will try it out when I get home.

If unzipping the attached file and invoking it by typing

seagate_check.sh

on the command line is a little too complicated, running seatools on your linux server, or flashing firmware on a hard drive is really going to tax your skills, regardless of the command involved.

 

Good luck...

 

Pull the drives out, read the label on the top of the drive.

 

Joe L.

 

Doh!

 

I thought it would have been more complicated then that!

 

Should I place the script in the root of my flash drive?

 

Thanks Joe!

Yes. Put it at the root of the flash drive, and then either cd to the folder first, and then invoke it

cd /boot

seagate_check.sh

 

or type the full path:

/boot/seagate_check.sh

 

As i mentioned, if smartctl does not run, or does not exist, you'll need to install it.

 

Another possibility to get the firmware version is to look in your syslog:

grep ATA /var/log/syslog

 

On my server you will find the model, serial, and firmware version among the matched lines.  It will be up to you to see if the model drive you have is in the potentially affected list.

Share this post


Link to post

I have 2 of the ST3100340AS (1T) drives.  They both use the "AD14" firmware, which does not require the firmware update.

Share this post


Link to post

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now