Maxtor 6Y160P0 disks: bad batch?

[Mark Mackey]

Hi all.

We've got a couple of disk servers with 3ware IDE-RAID cards in them.
The backup one of these just sits there and syncs with the main disk
server once a day, so it's not exactly heavily used.

This backup server, which is ~8 months old, has a 3ware 7506-8 connected
to 8 Maxtor 6Y160P0 160GB disks, and we keep a spare disk in a cupboard
to swap in in case of emergencies. Over the last 3-4 months, we've had
three disks die completely, and another two develop bad blocks. The
disks with bad blocks seem to respond to removing from the array,
reformatting with a bad-blocks check, and reinstating, but I'm not sure
how long they'll last.

Are the known problems with 6Y160P0s on 3ware cards, or with 6Y160P0s
generally? Having been bitten in the past by an IBM DeathStar, I'd like
to know if the 6Y160P0s are all crap or whether I've just ended up with
a bad batch. Has anyone else had any similar problems?

 

[Georg Acher]

<...>
|> Are the known problems with 6Y160P0s on 3ware cards, or with 6Y160P0s
|> generally? Having been bitten in the past by an IBM DeathStar, I'd like
|> to know if the 6Y160P0s are all crap or whether I've just ended up with
|> a bad batch. Has anyone else had any similar problems?

Yes, a 4 week old 250GB Maxtor died a horrible death. The reallocated sectors
reached the 2000 mark, before SMART called it "failing". Actually, it was the
first disk with problems I've ever seen where SMART really failed... It must be
some sort of head crash, it happened from one day to another on a heavily used
system. No chance of any data recovery.

Another 160GB Maxtor (about 5 months old) also suddenly showed increasing failing
sectors. The odd thing was that the disk wasn't able to reallocate the pending
sectors even when writing to them. The only chance to get the data back was to
copy the image to another disk and work on that image (took about two days...).
BTW: The backup on that system was switched off, because the boot disk, an older
Maxtor 30GB, got incredible slow: Just about 50KB/s, but absolutely no data
errors...

So, yes I think Maxtor has a slight quality problem, at least for the cheaper
disks. A 3ware SATA-Raid consisting of 6 250GB Maxtors has no reallocated sectors
at all after 8 months. I hope it will stay that way...

 

[General Schvantzkoph]

Hi all.

We've got a couple of disk servers with 3ware IDE-RAID cards in them.
The backup one of these just sits there and syncs with the main disk
server once a day, so it's not exactly heavily used.

This backup server, which is ~8 months old, has a 3ware 7506-8 connected
to 8 Maxtor 6Y160P0 160GB disks, and we keep a spare disk in a cupboard
to swap in in case of emergencies. Over the last 3-4 months, we've had
three disks die completely, and another two develop bad blocks. The
disks with bad blocks seem to respond to removing from the array,
reformatting with a bad-blocks check, and reinstating, but I'm not sure
how long they'll last.

Are the known problems with 6Y160P0s on 3ware cards, or with 6Y160P0s
generally? Having been bitten in the past by an IBM DeathStar, I'd like
to know if the 6Y160P0s are all crap or whether I've just ended up with
a bad batch. Has anyone else had any similar problems?

Maxtor drives are awful, in my experience they have a 50% out of the box
failure rate.

 

[Juhan Leemet]

<...>
|> Are the known problems with 6Y160P0s on 3ware cards...

Yes, a 4 week old 250GB Maxtor died a horrible death. The reallocated
sectors reached the 2000 mark, before SMART called it "failing"...

Was it making noise? Usually you can head a louder (than normal disks)
"zzzzzzz" as it's "ploughing a furrow" so to speak. I had a case of
"infant mortality" (crib death?) with a brand new Fujitsu disk. Made noise
right away, came up with lots of errors. I took it back to the retailer
and got an exchange right away, no questions asked: Fujitsu policy. That
sort of thing happens. I suspect somewhere along the line someone had
dropped that disk of mine? Other Fujitsu disks have been great!

Maxtors have worked well for me, but I haven't bought any high
capacity (my biggest is 60MB from years ago) recently. I've been picking
up some eBay SCSI disks for other systems. Lots of storage (NFS).

If Maxtor have "systematic" problems, that is a matter of some concern.

 

[Mark Mackey]

So, yes I think Maxtor has a slight quality problem, at least for the cheaper
disks. A 3ware SATA-Raid consisting of 6 250GB Maxtors has no reallocated sectors
at all after 8 months. I hope it will stay that way...

Our other disk server has been running non-stop for 2 years now on a set
of 8 6Y120P0s and they are all fine (no SMART errors on any of the
disks).

Yet another 6Y160P0 died this morning. That's 4 dead and two error-prone
disks out of a batch of 9. I'm not a happy bunny.

 

[Folkert Rienstra]

Our other disk server has been running non-stop for 2 years now on a set
of 8 6Y120P0s and they are all fine (no SMART errors on any of the
disks).

Did it occur to you that that might be because they are on "the other server"?

 

[james]

Maxtor drives are awful, in my experience they have a 50% out of the box
failure rate.

This so strange to me. I have 2 maxtor scsi hard drives and absolutely
love them. They are fast and quiet and I have never had a single
problem. I have had one of them for over a year, too. However, I hear so
many people complain, not only on the linux forums but mac and windows
forums too, how their maxtor drives are dying. Must be a difference
between the ide and scsi drives.

keep on computin'

jamess

 

[Juhan Leemet]

This so strange to me. I have 2 maxtor scsi hard drives and absolutely
love them. They are fast and quiet and I have never had a single
problem...I hear so many people complain...

It just might be because only unhappy people make noise? The only
positives would be like yours, in response to a "grumpy blast". I like my
Maxtor drive(s), too. As you say fast and quiet. Way they're supposed to be.

 

[Juergen Pfann]

This so strange to me. I have 2 maxtor scsi hard drives and absolutely
love them. They are fast and quiet and I have never had a single
problem. I have had one of them for over a year, too. However, I hear so
many people complain, not only on the linux forums but mac and windows
forums too, how their maxtor drives are dying. Must be a difference
between the ide and scsi drives.

Yes, I suppose so, too. Take into account, Maxtor "themselves" did not
manufacture any single SCSI HDD - before they bought Quantum, that is...

I'm very fond of my Quantum Atlas IV, too. However, a 40 GB Maxtor (54098U8)
is just dying, according to it's SMART status, and the 6Y80P0 had to be
RMA'd after 3 days or something. Even though the warranty replacement
seems to work (well, since some one month at least...), I am _not_
going to buy any Maxtor ATA HDD any more, I guess.

Juergen

 

[Mark Mackey]

Did it occur to you that that might be because they are on "the other server"?

It certainly did. All of the hardware sensors in both machines are
monitored, so there's no overheating problem (the disks stay at ~40C
when the machine's being hammered hard). The 3ware card works perfectly
as far as I can see, and I personally think it's a bit unlikely that a
subtle fault in the RAID card would cause physical damage to the
attached drives.

To follow up: Maxtor have now replaced all of the drives. We'll see
whether the new lot work OK.

 

[Folkert Rienstra]

Please setup your newsreader attribution line properly.
You are posting my reply address for everyone to see.

It certainly did. All of the hardware sensors in both machines are
monitored, so there's no overheating problem (the disks stay at ~40C
when the machine's being hammered hard). The 3ware card works perfectly
as far as I can see, and I personally think it's a bit unlikely that a
subtle fault in the RAID card would cause physical damage to the
attached drives.

Yes, but how about the Power Supply?

 

[Mark Mackey]

Please setup your newsreader attribution line properly.
You are posting my reply address for everyone to see.

It's the standard attribution line that comes with trn. In any case, if
your real email address is in your Reply-To field, then anybody scraping
email addresses off Usenet is going to grab it from there, no? Either
forge the whole thing (an in that case use example.com: that's what it's
for), or live with the results.

Yes, but how about the Power Supply?

I fail to see how a dodgy power supply could cause bad block errors on a
few different hard disks while leaving the rest of the system
unaffected.

 

[Folkert Rienstra]

It's the standard attribution line that comes with trn.

So?

It's not the standard attribution line for anything else as you can see
for yourself when you lookup other people's reactions to my posts.

Everything else uses the Sender address, not the Reply-To address.
It was pretty safe to use a Reply-To address until people like you
and Al Dykes came in here.

In any case, if
your real email address is in your Reply-To field, then anybody
scraping email addresses off Usenet is going to grab it from there,

No!

Either forge the whole thing (an in that case use example.com: that's what it's
for), or live with the results.

I won't. You are the one that is at fault, not me.

I fail to see how a dodgy power supply could cause bad block errors on a
few different hard disks while leaving the rest of the system unaffected.

Fine, keep killing your drives.

 

[General Schvantzkoph]

It's the standard attribution line that comes with trn. In any case, if
your real email address is in your Reply-To field, then anybody scraping
email addresses off Usenet is going to grab it from there, no? Either
forge the whole thing (an in that case use example.com: that's what it's
for), or live with the results.


I fail to see how a dodgy power supply could cause bad block errors on a
few different hard disks while leaving the rest of the system
unaffected.

Bad power connectors can especially if you have multiple drives on a
single power cable.

 

[Mark Mackey]

Bad power connectors can especially if you have multiple drives on a
single power cable.

That's worth knowing. In this case, however, the dead drives were pretty
randomly distributed over the array, so I doubt that a dodgy power cable
was involved. In any case, Maxtor have replaced the entire batch of
disks, so we'll see how the new lot go.

 

[Mark Mackey]

Everything else uses the Sender address, not the Reply-To address.
It was pretty safe to use a Reply-To address until people like you
and Al Dykes came in here.

Do you really think that if spammers were screenscraping addresses then
they'd pop up your message in Eudora, carefully copy down the contents
of the 'From:' field onto a piece of paper, then type it laboriously
into their spam address book? If it's in the headers, it's fair game:
why on earth do you think that having your email address in one header
but not another affords you any protection whatsoever?

I won't. You are the one that is at fault, not me.

Post the RFC, then.

Fine, keep killing your drives.

We'll see how the current batch from Maxtor do. It's worth noting that
the two disks which were swapped in very early on and which were from a
different batch to the others have had no problems at all.

 

[James]

Sorry to be replying kind of late.

Yes, I suppose so, too. Take into account, Maxtor "themselves" did not
manufacture any single SCSI HDD - before they bought Quantum, that is...

I thought that it was Quantum that bought Maxtor. My 18 GB Maxtor HDD
(oldest one) shows up in the SCSI list as a Maxtor HDD. My 36 GB Maxtor
HD (newest one) shows up as a Quantum. I don't know who made that 280 MB
Maxtor SCSI HDD that is in my old Apple IIGS, but it sure boots that
computer--still! The first SCSI HDD I bought for that Apple IIGS was a
Quantum; lasted about 1 year. It got a fatal problem called "stickum"
where some--super lubricant that was supposed to run real cool--turned
into cement after so many startups and shutdowns of the hard drive. I
guess one could honestly say that this was not really Quantum's fault
since the problem didn't crop up until about 1 year after the lubricant
was first put into use. That SCSI HDD was one of only two SCSI HDDS I
have had fail on me. The other was an IBM--talk about crappie HDDS!

I'm very fond of my Quantum Atlas IV, too. However, a 40 GB Maxtor (54098U8)
is just dying, according to it's SMART status, and the 6Y80P0 had to be
RMA'd after 3 days or something. Even though the warranty replacement
seems to work (well, since some one month at least...), I am _not_
going to buy any Maxtor ATA HDD any more, I guess.

So it could just be that the problem is that Quantum bought Maxtor and
turned their HDDs into crap. No one could blame you or anyone else for
not buying any more Maxtor ATA HDDs!

 

[John Turco]

Sorry to be replying kind of late.


I thought that it was Quantum that bought Maxtor. My 18 GB Maxtor HDD
(oldest one) shows up in the SCSI list as a Maxtor HDD. My 36 GB Maxtor
HD (newest one) shows up as a Quantum. I don't know who made that 280 MB
Maxtor SCSI HDD that is in my old Apple IIGS, but it sure boots that
computer--still! The first SCSI HDD I bought for that Apple IIGS was a
Quantum; lasted about 1 year. It got a fatal problem called "stickum"
where some--super lubricant that was supposed to run real cool--turned
into cement after so many startups and shutdowns of the hard drive. I
guess one could honestly say that this was not really Quantum's fault
since the problem didn't crop up until about 1 year after the lubricant
was first put into use. That SCSI HDD was one of only two SCSI HDDS I
have had fail on me. The other was an IBM--talk about crappie HDDS!

So it could just be that the problem is that Quantum bought Maxtor and
turned their HDDs into crap. No one could blame you or anyone else for
not buying any more Maxtor ATA HDDs!

Hello, James:

Actually, the reverse is true: Maxtor bought Quantum.


Cordially,
John Turco <[email protected]>

 

[J. Clarke]

Sorry to be replying kind of late.


I thought that it was Quantum that bought Maxtor.

Nope. Maxtor bought Quantum's hard disk operation--Quantum wanted to
concentrate on their tape business for some reason.

My 18 GB Maxtor HDD
(oldest one) shows up in the SCSI list as a Maxtor HDD. My 36 GB Maxtor
HD (newest one) shows up as a Quantum. I don't know who made that 280 MB
Maxtor SCSI HDD that is in my old Apple IIGS, but it sure boots that
computer--still! The first SCSI HDD I bought for that Apple IIGS was a
Quantum; lasted about 1 year. It got a fatal problem called "stickum"
where some--super lubricant that was supposed to run real cool--turned
into cement after so many startups and shutdowns of the hard drive.

That has nothing to do with quality. I've seen IBM, Seagate, and CDC drives
show the same behavior. And I've never seen any suggestion by any
knowledgeable person that CDC drives were not high quality.

I
guess one could honestly say that this was not really Quantum's fault
since the problem didn't crop up until about 1 year after the lubricant
was first put into use. That SCSI HDD was one of only two SCSI HDDS I
have had fail on me. The other was an IBM--talk about crappie HDDS!

Funny, IBM used to put those drives in their mainframes. Since mainframes
are leased and under a service contract any repair is a cost to IBM so they
tend to make them as reliable as they can.

So it could just be that the problem is that Quantum bought Maxtor and
turned their HDDs into crap. No one could blame you or anyone else for
not buying any more Maxtor ATA HDDs!

That would be true if Quantum had bought Maxtor.