Yes, but I assume that no matter how high the price will be for a permanent fix, it would always be a way lower than the data recovery itself. So, those drives can be like milking cows (or horses, your case )

_________________
Bitlevel Data Recovery

Horses and donkeys can be dangerous to milk , be sure you milk the RIGHT horse/donkey

(serious side note : donkey milk is good and has excellent anti-biothic properties , in case of otitis few drops of donkey milk stop the infection and heal it better and quicker than meds - it's a common known remedy here)

Dont you tell your customers that the disks are likely to fail again? Surely this is your responsibility to make sure that clients are aware that there is always a chance of additional failures.

I of course do as I cleared it in one of my previous posts. However, some peoples just stubbornly want to ignore all WARNINGS. So, what else can I do in such cases?

_________________
Bitlevel Data Recovery

- If you only reset S.M.A.R.T. (and translator), and do not upgrade firmware, the failure will come up again. I saw that once.
- If you apply a new firmware, the root cause for the failure is gone, so it will never happen again.

A question to all people, who sell new discs instead of giving back the repaired drive to customer: Did you ever think about the failure root cause, and failure probability / reliability? The behaviour was caused by a program failure. By installing the new firmware, the disc is fixed! Only if there are defective sectors, the disc must be replaced. Replacing in general is nonsense.
New technical products, e.g. hard discs, will fail much more often than discs, which were 1000-4000 hours in operation (my statistics regarding firmware issues).

Totally wrong. If you fix and update fw you don't have eliminated the root cause. I already said it.

The root cause for "busy" 7200.11 discs is a programming failure, which "forgets" to reset failure ring buffer pointer from 319 to 0. Firmware update will fix it. Am I sooo wrong?

Yes.

Here is the appropriate Seagate tech bulletin:

http://www.expreview.com/img/topic/seag ... Update.pdf

BTW, after you have been at this forum for a while, you will learn to ignore BlackST. His signal-to-noise ratio is essentially zero (that's absolute, not dB).

It's inversely proportional to your being nothing, that is near infinite.

_TK_ and fzabkar, perhaps you would then care to explain how this failure is seen in drives that do not have the "programming failure" (according to Seagate)

_________________
You don't have to backup all of your files, just the ones you want to keep.

Gents,

I don't like this polemic. If you believe, that the number of my posts is equivalent with the number of repaired drives - believe, what you want to believe.. Just check my last few posts in this forum, if you want.

I ever sent back the repaired drive with firmware upgrade to my customers, and only 1 drive came back. That was at an early time, when no firmware upgrade was available. The others are running fine until now. Of course, no drives with defective sectors are going back.

A lot of drives have not a 100% perfect firmware. Last week I repaired a WD3200AAKS, which was always in a busy state, in a similar way like all these seagate drives. And, of course some other firmware issues can happen. I know 5 of them, which I repaired by myself. Busy, translator failure, strange behaviour (e.g. klicking) at startup, Firmware "ZZ7L", "Firmware 1204F3C8".
Most of them were caused by external voltage drop. So the drive comes into an instable state, and blockes itself. Also dirty contacs to pre amp I have seen a lot of times. At WD and Seagate drives in a similar number.

BlackST, it would be interesting to know, if your drives have clean contacs, and your power supplies are in a good condition. How many blocked drives (absolute or percent) did you see after firmware upgrade?

Regards,

Thomas

I like polemics instead. But I don't accept the purpose of teasing and attack to obtain info , it's the old trick to lie to obtain truth . Play these dirty tricks elsewhere, maybe someone else will like them. Not me.

NORMAL MODE ON :

Stats : seen up today more than 100 mixed drives that "bricked" , with mixed FW like MX15, SD15 and many other. The very first were repaired but NO fw update was available and the problem started a lot of time ago so I don't know exactly what was their destiny. The others were fixed and almost all if possible FW updated as it was a waste to throw them away (but with a cautious caveat about the possibility of later failures) : invariably, the vast majority of them went into coma again so the FW update was useful as a tit on a bull (why people are complaining about the FW update and re-bricking ? wasn't it the panacea for all evil ? And what about the bricked drives that carry a SN or firmware that Seagate says NOT AFFECTED BY PROBLEM ?) Many of them are on controlled environment so I am sure about power, contacts and everything else. The problem was effectively elsewhere. The time between the fix and the re-bricking (in absence of countermeasures) was variable from 5-6 months to 1 year.
Bear in mind that 10-years (and even 5 nowadays) lifetime lasting drives are rare, are old drives or are specially designed drives that cost a lot.
The major flaw was in design and - as soon as you accept that they are cheap drives with many criticities and the update was not a definitive fix - if the purpose is to re-use a drive (complete refurb is yet to come) with some more reliability or simply squeeze out some extra months of life, you have to do something else.
ASSUMING THE DRIVE IS IN ACCEPTABLE SHAPE AND HEADS ARE WITHIN THE SPECS.
Otherwise, as usual, use it as a rifle target, boil them, eat them with oil and lemon, use as cannonball... .... ..... , crush them with a SUV or use them as doorstop or hi-tech paperweight. PERIOD.

P.S. @TK , I have tested voltage drop as it could be one cause, the worst thing happened were some drives that developed simple bads, some were automatically healed by reallocation, other needed a simple internal reformat then the surface was OK again. Wasn't able to brick intentionally at "user level" one of them, at least with my efforts. Oh well, if I really want to recreate the problem at higher level, I can

I'm sure, that a 100% perfect designed (and / or running) drive is not available. Also all mobile phones, computers and so on, which I had, don't have a perfect firmware. That's life. "God made machine language. All the rest ist the work of man."
If the owner of a disc accepts this simple truth and makes backup, almost everything would be ok. And - give him back the updated drive; he will make backups for sure!

External voltage drops or instable supply power, abeted by dried out electrolytic capacitors, most probably caused some of the strange firmware issues, which could not be easily solved by standard procedure. Also dirty contacts to pre amplifier (not external) caused in 2 cases wrong values in firmware area, which prevented drive to start. If you confuse drive in the moment, when it writes S.M.A.R.T. information or recalibrates, it will maybe start with failures. Not only seagate.

Until now I repaired ca. 70 Seagate drives with firmware problems, and after the firmware was updated, not one of them made problems again.
Every drive can fail, and will do that - sooner or later.


P.S.: Blocking the drive again? Should not be too difficult. Write back old firmware, and set pointer to the end. If I will have enough spare time, I'll try.

Maybe you were never contacted again and the drives failed again but this time the drive was discarded. Who knows ?

About more or less defective drives, I can ensure that there are much more reliable drives at a much more price. It's in any case a discussion that leads to nowhere : people usually want monster-storage drives for stacking mostly "porn and warez" - if someone stores critical irreplaceable data on 50$ drives with no "plan B" is simply irresponsible. Oh, I can't care less. I just know that usually, when knowing the price for recovery in case of major failure, data is always not important and... "you can keep the drive because I have to pay for disposal / I don't know where to bin it".
It's not a chance that on servers or critical application, drives are much smaller (there are high capacity drives too) and made very differently with very different firmware.
And the bad habit to put unfinished products on market because of the pressure from the management ("then we will fix it later with hotfix or FW update, don't worry") is hard to fight.

My eyes are ok, so you can use normal black letters.

I offered to all customers with these firmware issues free lifetime repair, if the firmware problem will come up again.

The main problem, and root cause for most of our daily work is a philosophic problem: customers and purchasing agents want to buy everything as cheap as possible. And the newest products (bigger, better, faster..) of course. Manufacturers of bulk goods must follow these requests, otherwise they will go bankrupped soon. That's real life, unfortunatelly.

Are you really sure, that e.g. server-drives (24/7) are more reliable then bulk drives? Of course they have a different firmware, which has stronger limits. But if I have a look into my database, I can't see a significant difference to bulk freight..

Are you really sure, that e.g. server-drives (24/7) are more reliable then bulk drives?

Yes. Otherwise there would be an ass-kicking chain

We are not talking about "el cheapo" drives on server. We are talking about 15K / 10K SCSI, UWSCSI, FC and SAS, or special series for SATA / ATA.
When they fail, usually it means a PITA for recovery.

Hmm, I have lots of them here (SCSI / UWSCSI, 10k, 15k). Defective of faulty.
15k / 10k drives are are high power drives. Electrical energy goes in, and will be transformed into heat. Thermal energy = aging.
FC or SAS drives are really good, but too expensive for private.

- Check the internet, how long e.g. WD Raptor will survive. After 3 years many of them are (almost) dead. (ok, that's 10x longer than 7200.11er Seagates with firmware bug)
- Lay e.g. a ST3500320AS and a ST3500320NS side by side and tell me, what's different (except firmware).
- Have a look into a S<x>N workstation. Inside: desktop drives.

Sometimes I wish to be something like god, and bring humans back to rationality. But if you see people in a supermarket, or driving on a street, you loose your faith. We live in a non-perfect world, therefore backup and RAID were invented.
(Sorry, a lot of polemic. Good night...!)

Don't shoot me, I'm just the messenger.

When the problem was first publicly acknowleged by Seagate not all drive models were included in the 'official list' of affected drives. As time went on the list grew. Maybe it is still growing!

Anyway I have my own opinion of what took place and members can laugh if they want but I truly believe Seagate leaked the lba 0 'dirty fix' in an attempt to limit their obligations and to keep their reputation from being stained. The public fix gets access to the data so end users have a chance if they follow the instructions religiously, but to say the firmware update fixes the problem is unclear.
320 power cycles might mean it could be 12 months or more before the original problem reappears

Maybe we should discuss..... 'Seagate busy/0 LBA problem twice on same drive after firmware update?' Has this actually happened? Is there any known case?