Drive Ready seek errors
Lately, a lot of my Linux boxes have had hard drive failures. There are about 15 servers running raid 1 on Red hat 7.3. Does anyone have any suggestions, maybe some good doc's, on hard drive troubleshooting / tuning? Some of the failed hard drives are only a couple months old, and have already been replaced twice, so I suspect a faulty ide controller. What is a good IDE controller card to buy for Linux?
Signature · 2 lines
Thanks,
Brandon E.Quoted 7 lines · Brandon Evans
Lately, a lot of my Linux boxes have had hard drive failures. There are about 15 servers running raid 1 on Red hat 7.3. Does anyone have any suggestions, maybe some good doc's, on hard drive troubleshooting / tuning? Some of the failed hard drives are only a couple months old, and have already been replaced twice, so I suspect a faulty ide
controller. What is a good IDE controller card to buy for Linux?
I can't answer the question about the controller cards, but have you considered using smartd? It should give you some warning that your drives
are on the way out. In addition, if this is happening across several
boxes, it is possible that kernel configuration is a problem. How do these drives fail? What kinds of errors do you get?
---
Toby Fisher Email: [email protected]
Tel.: +44(0)1480 417272 Mobile: +44(0)7974 363239
ICQ: #61744808
Quoted 7 lines · Toby Fisher
Quoted 1 line · Brandon Evans
Lately, a lot of my Linux boxes have had hard drive failures. There are about 15 servers running raid 1 on Red hat 7.3. Does anyone have any suggestions, maybe some good doc's, on hard drive troubleshooting / tuning? Some of the failed hard drives are only a couple months old, and have already been replaced twice, so I suspect a faulty ide controller. What is a good IDE controller card to buy for Linux?
I can't answer the question about the controller cards, but have you considered using smartd? It should give you some warning that your drives are on the way out. In addition, if this is happening across several boxes, it is possible that kernel configuration is a problem. How do these drives fail? What kinds of errors do you get?
Could it be a power suply problem affecting several machines?
Signature · 1 line
Pau Montero Pares
A message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 3 lines · Brandon Evans
Quoted 1 line
Could it be a power suply problem affecting several machines?
That's interesting because we recently had a power issue at our data center.
Can you verify if the hard disk armageddon happened just after this
power issue?
Signature · 1 line
Pau Montero Pares
Quoted 9 lines · Pau Montero Pares
Quoted 4 lines · Brandon Evans
Quoted 1 line
Could it be a power suply problem affecting several machines?
That's interesting because we recently had a power issue at our data center.
Can you verify if the hard disk armageddon happened just after this
power issue?
We had a spat of HD failures a few years ago that coincided with a period of low and fluctuating line voltage.
CD
Quoted 11 lines · Pau Montero Pares
Quoted 4 lines · Brandon Evans
Quoted 1 line
Could it be a power suply problem affecting several machines?
That's interesting because we recently had a power issue at our data center.
Can you verify if the hard disk armageddon happened just after this
power issue?
I don't think it was the power issue. After talking with a co-worker I found out The power problem was due to a miscommunication. he thought we had X amount of amps, but had less. So when the last server got pluged in, the fuse went. plus, the raq with the power problem was on a separate circuit then the raq that has the servers with the hard drive problems.
Signature · 2 lines
Thanks,
Brandon E.Quoted 6 lines · Toby Fisher
Quoted 1 line · Brandon Evans
Lately, a lot of my Linux boxes have had hard drive failures. There are about 15 servers running raid 1 on Red hat 7.3. Does anyone have any suggestions, maybe some good doc's, on hard drive troubleshooting / tuning? Some of the failed hard drives are only a couple months old, and have already been replaced twice, so I suspect a faulty ide controller. What is a good IDE controller card to buy for Linux?
Quoted 7 lines
I can't answer the question about the controller cards, but have you considered using smartd? It should give you some warning that your drives are on the way out. In addition, if this is happening across several boxes, it is possible that kernel configuration is a problem. How do these drives fail? What kinds of errors do you get?
---
Toby Fisher Email: [email protected]
Tel.: +44(0)1480 417272 Mobile: +44(0)7974 363239
ICQ: #61744808
-
I havent used smartd yet, but I will look into it. The kernel configuration could be a problem, since redhat stoped supporting 7.3, I have been building our own for security reasons. I'll search around the kernel archive.
Here is the error log from one of the servers.
Jun 2 17:31:20 server1 kernel: ide0: reset: success
Jun 2 17:31:21 server1 kernel: hda: write_intr error1: nr_sectors=1, stat=0x51
Jun 2 17:31:21 server1 kernel: hda: write_intr: status=0x51 { DriveReady SeekComplete Error }
Jun 2 17:31:21 server1 kernel: hda: write_intr: error=0x10 { SectorIdNotFound }, LBAsect=234436369, sector=232219399
Jun 2 17:31:21 server1 kernel: hda: write_intr error1: nr_sectors=1, stat=0x51
Jun 2 17:31:21 server1 kernel: hda: write_intr: status=0x51 { DriveReady SeekComplete Error }
Jun 2 17:31:21 server1 kernel: hda: write_intr: error=0x10 { SectorIdNotFound }, LBAsect=234436369, sector=232219399
Jun 2 17:31:21 server1 kernel: hda: write_intr error1: nr_sectors=1, stat=0x51
Jun 2 17:31:22 server1 kernel: hda: write_intr: status=0x51 { DriveReady SeekComplete Error }
Jun 2 17:31:22 server1 kernel: hda: write_intr: error=0x10 { SectorIdNotFound }, LBAsecttr error1: nr_sectors=1, stat=0x51
Jun 2 17:31:22 server1 kernel: hda: write_intr: status=0x51 { DriveReady SeekComplete Error }
Jun 2 17:31:22 server1 kernel: hda: write_intr: error=0x10 { SectorIdNotFound }, LBAsect=234436369, sector=232219399
Jun 2 17:31:22 server1 kernel: hda: write_intr error1: nr_sectors=1, stat=0x51
Jun 2 17:31:22 server1 kernel: hda: write_intr: status=0x51 { DriveReady SeekComplete Error }
Jun 2 17:31:22 server1 kernel: hda: write_intr: error=0x10 { SectorIdNotFound }, LBAsect=234436369, sector=232219399
Jun 2 17:31:22 server1 kernel: ide0: reset: successSignature · 2 lines
Thanks,
Brandon E.Quoted 25 lines · Brandon Evans
Here is the error log from one of the servers.
Jun 2 17:31:20 server1 kernel: ide0: reset: success Jun 2 17:31:21 server1 kernel: hda: write_intr error1: nr_sectors=1, stat=0x51 Jun 2 17:31:21 server1 kernel: hda: write_intr: status=0x51 { DriveReady SeekComplete Error } Jun 2 17:31:21 server1 kernel: hda: write_intr: error=0x10 { SectorIdNotFound }, LBAsect=234436369, sector=232219399 Jun 2 17:31:21 server1 kernel: hda: write_intr error1: nr_sectors=1, stat=0x51 Jun 2 17:31:21 server1 kernel: hda: write_intr: status=0x51 { DriveReady SeekComplete Error } Jun 2 17:31:21 server1 kernel: hda: write_intr: error=0x10 { SectorIdNotFound }, LBAsect=234436369, sector=232219399 Jun 2 17:31:21 server1 kernel: hda: write_intr error1: nr_sectors=1, stat=0x51 Jun 2 17:31:22 server1 kernel: hda: write_intr: status=0x51 { DriveReady SeekComplete Error } Jun 2 17:31:22 server1 kernel: hda: write_intr: error=0x10 { SectorIdNotFound }, LBAsecttr error1: nr_sectors=1, stat=0x51 Jun 2 17:31:22 server1 kernel: hda: write_intr: status=0x51 { DriveReady SeekComplete Error } Jun 2 17:31:22 server1 kernel: hda: write_intr: error=0x10 {
Quoted 8 lines
SectorIdNotFound }, LBAsect=234436369, sector=232219399
Jun 2 17:31:22 server1 kernel: hda: write_intr error1: nr_sectors=1, stat=0x51 Jun 2 17:31:22 server1 kernel: hda: write_intr: status=0x51 { DriveReady SeekComplete Error } Jun 2 17:31:22 server1 kernel: hda: write_intr: error=0x10 { SectorIdNotFound }, LBAsect=234436369, sector=232219399 Jun 2 17:31:22 server1 kernel: ide0: reset: success
Ok, I had similar problems, which i can only guess were controller-related as I have 1 20 gb ibm gxp drive and 1 40 gb Segate Baracuda drive.
In the kernel configuration, you should find the option:
Include IDE/ATA-2 DISK support
Immediately below this, you should find:
Use multi-mode by default
YHou should say yes to this. if you read the help, you will see one of you errors mentionned.
HTH
Signature · 3 lines
Toby Fisher Email: [email protected] Tel.: +44(0)1480 417272 Mobile: +44(0)7974 363239 ICQ: #61744808
Oh yes, one other thing. You'll probably notice some errors in dmestg in your boot-up, but after this things should settle down, they did for me in any case.
Cheers.
Signature · 3 lines
Toby Fisher Email: [email protected] Tel.: +44(0)1480 417272 Mobile: +44(0)7974 363239 ICQ: #61744808
on "6-2-2004" "Brandon Evans" writ:
: Lately, a lot of my Linux boxes have had hard drive failures.
... ciao:
in the not too distant past, word of 'faulty' 20 and 40GB drives
surfaced. 'maxtor' was one of the names mentioned, but was not alone in this issue ...
Signature · 2 lines
... i'm a man, but i can change,
if i have to , i guess ...Quoted 11 lines · terry white
on "6-2-2004" "Brandon Evans" writ:
: Lately, a lot of my Linux boxes have had hard drive failures.
... ciao:
in the not too distant past, word of 'faulty' 20 and 40GB drives
surfaced. 'maxtor' was one of the names mentioned, but was not alone in this issue ...
These are all 120 Gig hardrives, some Seagate, some Maxtor, and some western digital, So I do not think it is any one manufacturer at fault.
Signature · 2 lines
Thanks,
Brandon E.0
Quoted 18 lines
Quoted 11 lines · terry white
on "6-2-2004" "Brandon Evans" writ:
: Lately, a lot of my Linux boxes have had hard drive failures.
... ciao:
in the not too distant past, word of 'faulty' 20 and 40GB drives
surfaced. 'maxtor' was one of the names mentioned, but was not alone in this issue ...These are all 120 Gig hardrives, some Seagate, some Maxtor, and some western digital,
So I do not think it is any one manufacturer at fault.
Quoted 8 lines
--
Thanks,
Brandon E.
-
To unsubscribe from this list: send the line "unsubscribe linux-admin" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
I have experience with Adaptec 2410SA (sata) (built in driver in 2.4.x kernels)
Promise FastTrak 100 TX4 (ide) (kernel module from promise.com) FastTrak S150 TX4 (sata) (kernel module from promise.com)
The boxes wich have the adaptec controller are statisticly the most stable (no hdd crash yet) till now BUT these boxes also have the best COOLING system from the others that use the other raid controllers.Cooling is a very important constant that you must have in mind when building boxes with raid arrays that do heavy work (production servers).I get disk failures all the time because of the bad cooling system.
Also remember to enable all the loging options in the scsi options while configuring your kernel.
Signature · 2 lines
Alexander Economou GNET NOC
Re: a Maxtor story (was: Drive Ready seek errors)
In reply to Brandon Evans-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Quoted 2 lines · Brandon Evans
These are all 120 Gig hardrives, some Seagate, some Maxtor, and some western digital,
I've had nothing but trouble with Maxtor's Diamond Plus series. A few months ago I bought 3 120Gig drives and set up a linux box which worked like a charm. One day I noticed that a partition located within the 30-40% of cylinders of one drive was giving bad data read results, about 1-2 MB/s[1]. So I ran a full fsck which took a long time but worked out fine. Over time, the other drives showed exactly the same symptoms.
Since areas where bad data read results appeared were spreading I first used the smart daemon tools to check if the drives were failing. I haven't got the logs anymore but it was suggested to run a full diagnostic. So I did with Maxtor's PowerMAX utility.
I got the diagnostic code de699c79 along with a message that the hard drive is failing. At first the code was not accepted at Maxtor's online warranty site located at
http://www.maxtor.com/de/support/service/warranty_ata.htm (it is by now, btw) so I contacted the support. It was suggested to low level format the drive and I did as suggested but running PowerMAX afterwards still gave the same diagnostic code.
In the end no data was lost (due to a paranoid backup strategy) and all Maxtor drives were replaced by WD ones. For me, buying Maxtor drives is considered A Bad Thing.
[1] I used the same partition layout for all 3 drives (with ext3 fs), no RAID whatsoever, so it was easy to track and locate the problem. FWIW, the server is/was used for several large MySQL databases and raw analytical data of some spectrometers.
- --
Bastard Administrator in $hell
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
iD8DBQFAwZ1eLMyTO8Kj/uQRAnXHAJ445wNrrzimKcwynyYizKIeAbS8zQCeJvx5 tCznuYNm70bZlD0lu9THRT0=
=eyo3
-----END PGP SIGNATURE-----
Re: a Maxtor story
Unfortunately it is not just the 120Gb Maxtor hard drives - over a period of about 12 months I bought 22 Maxtor hard drives, of these 14 have now failed. Almost exactly the same problem as markus, In my case, extreme slowdown - followed by sudden failure.
The majority of these disks were the slimline models (never again!), but there was at least one normal(?) sized hard drive. The capacity ranged from 30 Gb to 120Gb
Needless to say, after replacing the lot with IBM deskstars, I have not had a single disk failure (to date) and will not be buying Maxtor disks again.
Thankfully, I also have a highly paranoid backup system - and did not lose any data, and am now running the smart daemon tools on all the critical machines.
Just my 2 pence
Ben Potter
Quoted 18 lines · markus reichelt
I've had nothing but trouble with Maxtor's Diamond Plus series. A few months ago I bought 3 120Gig drives and set up a linux box which worked like a charm. One day I noticed that a partition located within the 30-40% of cylinders of one drive was giving bad data read results, about 1-2 MB/s[1]. So I ran a full fsck which took a long time but worked out fine. Over time, the other drives showed exactly the same symptoms.
Since areas where bad data read results appeared were spreading I first used the smart daemon tools to check if the drives were failing. I haven't got the logs anymore but it was suggested to run a full diagnostic. So I did with Maxtor's PowerMAX utility.
In the end no data was lost (due to a paranoid backup strategy) and all Maxtor drives were replaced by WD ones. For me, buying Maxtor drives is considered A Bad Thing.
Re: a Maxtor story
Hello,
I have a Maxtor 6E040L0 40G and it's working perfectly. You guys are getting me worried...what's the medium time until failure? I have it for almost a year now and never had problems with it.
Quoted 14 lines · Ben Potter
Unfortunately it is not just the 120Gb Maxtor hard drives - over a period of about 12 months I bought 22 Maxtor hard drives, of these 14 have now failed. Almost exactly the same problem as markus, In my case, extreme slowdown - followed by sudden failure.
The majority of these disks were the slimline models (never again!), but there was at least one normal(?) sized hard drive. The capacity ranged from 30 Gb to 120Gb
Needless to say, after replacing the lot with IBM deskstars, I have not had a single disk failure (to date) and will not be buying Maxtor disks again.
Quoted 6 lines
Thankfully, I also have a highly paranoid backup system - and did not lose any data, and am now running the smart daemon tools on all the critical machines.
Just my 2 pence
Ben Potter
---------------------------------------------------------------
Incearca acum noul sistem de dating oferit de portalul acasa.ro
Re: a Maxtor story
For me, the average failure rate was around 5 months - however I have 2 maxtors at home (1 80Gb and 1 60Gb - need to check model) that have been running for a couple of years without a single glitch.
It looks (to my highly inexperienced eye) as if maxtor have had a faulty batch.
I wouldn't worry about it, unless you start to notice a considerable slowdown
Alternatively, try running the smart daemon tools, on checking all of mine showed up with a lot of warnings - but my bios hadn't reported them (probably my fault)
I now periodically check all hard drives (48 of them) with the smart tools (roughly once a month) - it takes a bit of time - but it is better than one failing in the middle of the day and a user losing work (and screaming at me)
Hope this helps
Ben Potter
Quoted 28 lines · DragonK
Hello,
I have a Maxtor 6E040L0 40G and it's working perfectly. You guys are getting me worried...what's the medium time until failure? I have it for almost a year now and never had problems with it.
Quoted 11 lines · Ben Potter
Unfortunately it is not just the 120Gb Maxtor hard drives - over a period of about 12 months I bought 22 Maxtor hard drives, of these 14 have now failed. Almost exactly the same problem as markus, In my case, extreme slowdown - followed by sudden failure.
The majority of these disks were the slimline models (never again!), but there was at least one normal(?) sized hard drive. The capacity ranged from 30 Gb to 120Gb
Needless to say, after replacing the lot with IBM deskstars, I have not had a single disk failure (to date) and will not be buying Maxtor disks again.
Thankfully, I also have a highly paranoid backup system - and did not lose any data, and am now running the smart daemon tools on all the critical machines.
Just my 2 pence
Ben Potter---------------------------------------------------------------
Incearca acum noul sistem de dating oferit de portalul acasa.ro-
To unsubscribe from this list: send the line "unsubscribe linux-admin" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: a Maxtor story
You guys got me worried too. We have a 16-disk RAID 5 array, it's all Maxtor 250G SATA drive (the company supply the array recommand the Maxtor, since they have the lowerest failure time on ATA), and it's been working good for 9 months now, and we do parity-regenaration each months, found one or two bad blocks which had been remapped after. I will see how this thing goes.
Thanks for the inputs!
Chen
Quoted 73 lines · Ben Potter
For me, the average failure rate was around 5 months - however I have 2 maxtors at home (1 80Gb and 1 60Gb - need to check model) that have been running for a couple of years without a single glitch.
It looks (to my highly inexperienced eye) as if maxtor have had a faulty batch.
I wouldn't worry about it, unless you start to notice a considerable slowdown
Alternatively, try running the smart daemon tools, on checking all of mine showed up with a lot of warnings - but my bios hadn't reported them (probably my fault)
I now periodically check all hard drives (48 of them) with the smart tools (roughly once a month) - it takes a bit of time - but it is better than one failing in the middle of the day and a user losing work (and screaming at me)
Hope this helps
Ben PotterQuoted 33 lines · DragonK
Hello,
I have a Maxtor 6E040L0 40G and it's working perfectly. You guys are getting me worried...what's the medium time until failure? I have it for almost a year now and never had problems with it.
Quoted 14 lines · Ben Potter
Unfortunately it is not just the 120Gb Maxtor hard drives - over a period of about 12 months I bought 22 Maxtor hard drives, of these 14 have now failed. Almost exactly the same problem as markus, In my case, extreme slowdown - followed by sudden failure.
The majority of these disks were the slimline models (never again!), but there was at least one normal(?) sized hard drive. The capacity ranged from 30 Gb to 120Gb
Needless to say, after replacing the lot with IBM deskstars, I have not had a single disk failure (to date) and will not be buying Maxtor disks again.
Quoted 6 lines
Thankfully, I also have a highly paranoid backup system - and did not lose any data, and am now running the smart daemon tools on all the critical machines.
Just my 2 pence
Ben PotterQuoted 7 lines
---------------------------------------------------------------
Incearca acum noul sistem de dating oferit de portalul acasa.ro-
To unsubscribe from this list: send the line "unsubscribe linux-admin" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html-
To unsubscribe from this list: send the line "unsubscribe linux-admin" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Signature · 13 lines
=========================================== Yu Chen Howard Hughes Medical Institute Chemistry Building, Rm 182 University of Maryland at Baltimore County 1000 Hilltop Circle Baltimore, MD 21250 phone: (410)455-6347 (primary) (410)455-2718 (secondary) fax: (410)455-1174 email: [email protected] ===========================================
Re: a Maxtor story
In reply to DragonKQuoted 9 lines · DragonK
Hello,
I have a Maxtor 6E040L0 40G and it's working perfectly. You guys are getting me worried...what's the medium time until failure? I have it for almost a year now and never had problems with it.
On Sat, 05 Jun 2004 17:10:30 +0100, Ben Potter
<[email protected]> wrote:
Maybe you got lucky. I have purchased many of the same in the last year and have returned at least 60% of them, all within 3-6 months of activation; some only weeks. The RMA's were getting so bad, I even had drives coming back to me with Linux still on from the last person that set it back. What's up with that?
Seems even IBM has moved away from installing them in their desktop machines.
IMO Maxter has gone the way of Western Digital, sucketh.
Since I moved to the slightly more expensive Seagate Barracuda 7200.7 drive, I have had much less hardware problems. Of course I would never put anything less then IBM 15,000 RPM SCSI drives in a server. =P
--
Scott
Quoted 4 lines · Brandon Evans
Lately, a lot of my Linux boxes have had hard drive failures. There are about 15 servers running raid 1 on Red hat 7.3. Does anyone have any suggestions, maybe some good doc's, on hard drive troubleshooting / tuning? Some of the failed hard drives are only a couple months old, and have already been replaced twice, so I suspect a faulty ide controller. What is a good IDE controller card to buy for Linux?
the best thing is to follow the following approach:
+ to download a disk utility from the manufacturer to check the integrity at low level of your hard disk drive.
+ after discarding this possibility, go ahead with fsck checks on your hard disk drives (linux checks).
+ if both of these checks are negative (no problem at all), then you should start checking something with your configuration of your RAID controller.
Anyway, googling a little bit about your Hard disk drive model and RAID controller could give you a few clues if there have been problems with that configuration in the past...
BR,
Miguel
-
To unsubscribe from this list: send the line "unsubscribe linux-admin" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html