Sign in to follow this  
blueznl

'lemon' 6Y080L0 sata

Recommended Posts

this can't be for real, i told myself, so i investigated a little more... i borrowed another 80 gb maxtor and checked it out...

this is *very* weird... perhaps some other expert can shine their light on it, because i haven't got a clue... i know how to reproduce it, i just don't know why it happes (though i stronly suspect the drives

read on...

1. configuration: asus k8n-e, updated bios and drivers, windows xp, 3 harddisks, hd1 and hd2: maxtor 'lemon' 6Y080L0, hd3: maxtor 6Y120P0, windows xp pro sp2

2. when copying a 'large' amount of data (bunch of 200 mb avi files) from hd3 to hd3 (pata) transfer speed is initially 23 MBps, then slows down to 17 or so MBps, this is quite normal behaviour

3. when copying the same files from hd1 to hd1 OR hd2 to hd2 transfer speed starts at 17 MBps then slows down to 2.3 MBps (!), you can hear the drive heads make additional movements

4. now comes the interesting part: when copying this set of files from hd1 to hd1 AND AT THE SAME TIME copy a similar set ON ANY OTHER DRIVE (ie. hd1 to hd1 or hd2 to hd2) suddenly the speed PICKS UP again

5. it looks like this behaviour only happens when you do a 'disk to disk' copy action on the same sata drive, it does not happen between different drives, it does not happen when simultaneously other actions (even entirely restricted to other drives) take place, it does not happen on pata drives at all

6. i've repeated the above with an asrock mainboard with an xp2500 cpu and a sil 3114 sata controller (so an entirely different hardware setup with only the same sata drives in comon) and it showed the same behaviour

7. so i tried something even more stupid: two copy actions on the same drive at the same time, id. hd1 to hd1 and another hd1 to hd1... one copying thread ran from 2 MBps to 4 MBps, two threads (copying different stuff of similar sizes) at the same time totalled higher 8 to 10 MBps

8. interestingly, this does not happen immediately, but takes a few seconds before drive speed starts slowing down, i wondered if it could be the power supply or something similar, but doing multiple copy actions on multiple drives did not cause this problem, even though that would be using more power

9. this strange behaviour also showed up when doing raid0... writing speed was awfully slow, even though reading speed was all the way up where it should be

10. could it be windows? unlikely, as it doesn't show up on pata... could it be the mainboard / chipset? unlikely, as it shows up on another mainboard / chipset as well... could it be a single bad drive? unlikely, as i've seen it on more than one drive... interference and thus data loss on the cable? unlikely, as two copy actions on the same drive result in a higher net result...

very weird, this, i am tempted to think that the (caching?) logic on board the 6Y060L0 is not up to the job, perhaps some sort of protection logic kicks in too early? i dunno but i can easily reproduce this situation with these drives

i would like to know if other people have similar experience with sata drives, it's easy to test: just copy one directory with many large files to another directory on the same drive

Share this post


Link to post
Share on other sites

It may just be normal windows behaviour. Only way to tell is to get 2 different hard drives and see if the performance improves.

Behaviours 2 & 3 seem like what I’d expect. I’m not sure why performance improved when running multiple file system threads at the same time, but it could just be windows.

Share this post


Link to post
Share on other sites
Maxtors have all kinds of weird problems (also contributing to its unusually high failure rate). Check this out:

http://www.dataclinic.co.uk/maxtor-hard-disk-recovery.htm

On one of the lines in that page: " * run slowly" .

Moral is blame Maxtor! :angry:

197470[/snapback]

Well, if you look at the main site of dataclinic, they have list of fujitsu and western digital as well. Beside, I don't think quoting a data recovery company for reliability of HD in the past is a good indicator, they may only show what is their experties in.

from what i know the D740, 540X, and Fireball 3 has no reliability problem what so ever, so why would they put it there? (maybe they were a maxtor warranty contractor in the past).

Share this post


Link to post
Share on other sites
It may just be normal windows behaviour. Only way to tell is to get 2 different hard drives and see if the performance improves.

Behaviours 2 & 3 seem like what I’d expect. I’m not sure why performance improved when running multiple file system threads at the same time, but it could just be windows.

197477[/snapback]

Check what UDMA mode you are running in and whether you can swap the IDE cable, I had the same problem in the past running 40pin cable and running in PIO mode because I want to save a few penny. Check that one first.

Share this post


Link to post
Share on other sites
Well, if you look at the main site of dataclinic, they have list of fujitsu and western digital as well. Beside, I don't think quoting a data recovery company for reliability of HD in the past is a good indicator, they may only show what is their experties in.

from what i know the D740, 540X, and Fireball 3 has no reliability problem what so ever,  so why would they put it there? (maybe they were a maxtor warranty contractor in the past).

197504[/snapback]

Did you checkout the PDF documents in that link I gave? They explain how the drives failed in detail. ;)

Share this post


Link to post
Share on other sites

Diamondmax Plus 9 failure rate is in line with other brands, at least IME. We've been through a couple hundred in the last month or so and failure rate looks pretty normal to us.

Share this post


Link to post
Share on other sites

problem solved

there appears to be a firmware issue with the 80 gb diamondmax sata drives, using another drive brand solved all problems

i guess people should be warned against the use of the 6Y080L0 sata drive, it appears to have two shortcomings:

1. it slows down during extensive / prolongued read / write sessions on a single drive

2. it performs bad in raid 0 configurations (slow writing, very fast reading though)

note: regular users may not notice this behaviour, as it only happens when the caches of the drives get saturated

the shop where i bought my set just had a second batch being swapped with the same issue, so fortunately i wasn't the only person experiencing this!

my solution? i fooled around a little more with raid, found out that raid on an asus k8n-e is not a good idea, then settled for a new raptor wd740 as my boot device... now when windows crashes next time, at least it crashes fast :-)

Share this post


Link to post
Share on other sites
there appears to be a firmware issue with the 80 gb diamondmax sata drives, using another drive brand solved all problems

As confirmed, its another firmware issue (as stated in my links).. maybe Maxtor should recall these drives and issue a new firmware. ;)

Share this post


Link to post
Share on other sites

hey lexwalker, any company that specializes in data recovery is going to claim there are some issues with any drive, so that is hardly a reference :-)

and to maxtor's defense: at least the drives lasted for more than a week, i just had a wd740 crashing on me inside 4 days...

Share this post


Link to post
Share on other sites
hey lexwalker, any company that specializes in data recovery is going to claim there are some issues with any drive, so that is hardly a reference :-)

and to maxtor's defense: at least the drives lasted for more than a week, i just had a wd740 crashing on me inside 4 days...

4 drives failed within 4 days? That's bad... :o

Are those drives all inside a single PC? Its possible the PSU could not supply enough power to all 4 drives at once, hence causing those failures. A faster drive like Raptor consumes more power than usual, multiply that by 4 and definitely more power is needed. ;) Maybe try using PSU brands like Enlight, Acbel, Enermax, etc... (I'm guessing only here, but sounds like the PSU may be the cause of your "unexpected" drive failures..)

Btw, I have 4 internal drives here, 1 CDRW and 1 DVDRW.. (plus 4 external drives on USB) so far I hadn't had any failed on me in the past 2 to 3 years. However I changed my PSU to an Enlight 420W "true power" unit (sold the original "unknown brand" PSU away). My previous rig (P3 machine) also use an Enlight 300W one and that one served me plus 4 hard drives without problems (still kept that PSU for emergency use :lol: ).

Ps. all my drives are Seagates...

Edited by lexwalker

Share this post


Link to post
Share on other sites

> Are those drives all inside a single PC?

euh, you didn't read well :-) it's two maxtors that had a firmware issue, and the wd that crashed inside 4 days :-)

obviously, i have 80 pins cables etc. don't worry, i'm not *entirely* a beginner :-)

> Its possible the PSU could not supply enough power to all 4 drives at once

>, hence causing those failures. A faster drive like Raptor consumes more power

3 drives, one sata wd740 raptor, one sata wd2000jb, and one pata maxtor 6y120l0 pata (hope i got that last letter right, anyway the pata 8 mb cache version)

> than usual, multiply that by 4 and definitely more power is needed. ;)

> Maybe try using PSU brands like Enlight, Acbel, Enermax, etc... (I'm guessing

> only here, but sounds like the PSU may be the cause of your "unexpected"

> drive failures..)

got a good quality dual fan 450W 'silentmaxx', german brand, true power, power correcting etc. (yeah right, perhaps a german *label*) iirc, just to be sure, got some okay reviews

Share this post


Link to post
Share on other sites
From your description, your hardware looks fine.

Did you manage to solve the problems?

What's your working RAID HDD configuration at the moment?

:rolleyes:

198431[/snapback]

it's working now, yep, follow the link

i tested a little more, and decided to use a wd740 as single boot drive, performance win was not sufficient to warrant the use of a raid set (writing played too much a role, i'm using the machine mostly for development, and compiler runs don't like write delays :-))

Share this post


Link to post
Share on other sites

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
Sign in to follow this