Partitioning large HD under XP

[djc]

Note that if your new disk is SATA rather than ATA/IDE it will need XP complete with SP2 on the install disk in order to recognise a SATA drive.

[John Rumm]

It is probably also worth remembering that most manufacturers have turned out a range that has proved to be a turkey from time to time. IBM never really recovered from the Deathstar incident, and WD had a rang3e of very dodgy Caviar drives for a while.

[Daniel James]

The fact that it was manufactured post-Seagate doesn't necessarily mean that any problems in Maxtor's design (it may have been designed pre-Seagate, remember) or manufacturing process have been rectified. One would like to think that there will have been some improvement, but Seagate may have bought Maxtor for any number of reasons, and may just be letting them do their own thing.

Maxtor drives are almost always cheaper than the equivalent Seagate, which should suggest that their drives are not identical ... whether they are of comparable quality is another matter.

The trouble with the hard drive business is that one never really has good information as to which drives make a good purchase until after those drives have been superseded and are no longer available, as that's when their reputation for reliability (or otherwise) can be established beyond any doubt.

Cheers, Daniel.

[Daniel James]

Small, no. Statistically significant -- given the millions of drives that Maxtor must make every year -- probably not.

I would say that you've either been unlucky, or that you have been using the drives outside their ideal operating conditions, or that your supplier drop-tests disks onto concrete before shipping ... or something else.

I can't believe that 17.5% of all Maxtors fail in service ... that would be headline news.

Cheers, Daniel.

[Andy Hall]

There were multiple batches from different suppliers over a period of nearly a year.

The drives were fitted in equipment used on a clean power supply in air conditioned data centres.

They may not. All I can say is that this proportion of units failed out of a range that I bought over a period of time.

It appears that I am not the only person to report poor reliability with this brand.

[Dave Baker]

I think life's too short to worry about every detail of whether you've bought the best hard drive you could have done with the money. I generally just go to computerprices.co.uk and buy the cheapest drive of the capacity I'm wanting. Some are going to fail early whatever make you choose but the chances are probably fairly low. My Maxtor 250 gb 7Y250PO was bought in September 2005 and has been faultless so far. It works very hard as the pc is on 24/7 downloading several gb of files per day. I think you can safely say that barring holidays it's been reading and writing to disk continuously for 2 years. Maybe that's better for them than being switched on and off all the time. I really don't know.

What I find more extraordinary is not that the occasional disk fails but how we take for granted that everything gets faster, bigger but also cheaper year after year. In 1987 when I bought my Amstrad 1512 a 30mb hard card for it cost £300. My 250gb Maxtor cost £80 and now a 500gb disk is about £60. No doubt in two more years time a 1000gb disk will be less still.

[Van Helsing]

I went through a period of having problems with Maxtor's threw a couple away and got a couple replaced under warranty.

When the last one started to gives errors I switched it to an old system just to run through the Seagate/Maxtor diagnostics & low level format. It passed the diagnotic with no problem and after the reformat has been fine, although in a different system.

After a bit of Googling came across more than one suggestion that Maxtors could be picky about the IDE controller they're connected to. On the surface this seems to support my experiences.

VH.

[The Natural Philosopher]

It is, around te sorts of places where people care.

[Daniel James]

That's not necessarily the kindest environment to be running the disks in. In a data centre -- where the racks generate so much heat that you need to air-condition the building -- the temperature *inside* the racks themselves can often be quite high.

Although Maxtor state a maximum operating temperature for their drives that is higher than that of many other manufacturers there have been reports that their reliability actually suffers *more* at elevated temperatures than that of other makes.

See:

words on the matter that should carry more weight than mine.

Were you using disks rated for data-centre use, or ordinary cheap consumer drives made for home PCs? Disks rated for 24/7 operation can cost up to twice as much, but the hassle of continually having to change the drives your racks (and restore from backups/recreate the RAID array) must surely make it worth the cost... If you are using cheapies you shouldn't really be surprised that they sometimes fail.

Above all, though: keep those drives cool.

Cheers, Daniel.

[Huge]

I think you need to read Google's recently published study on disk lifetimes in their data centres;

[Andy Hall]

It could be, but it wasn't. Each rack has individual forced air cooling and temperature monitoring every 4U.

When drives started to fail, thermocouples were installed inside the chassis next to the drives.

These were meant to be the quality products

Yes I know.

I never use cheap anything.

Obviously.

The reality was that the products were used well within specification and they had a much higher than reasonable failure rate.

All were replaced with Seagate. Zero failures since.

[Daniel James]

I have read it. I really don't think it says anything terribly useful, given that Google didn't specify the brands or models of disks that they used, or provide any breakdown of -- or correlation between -- the failures they observed by brand/model, age of drive, operating environment (temperature), or failure mode.

There are a lot of woolly hand-waving statements, with no clear figures to back them up, and too little supporting data for it to be possible to understand clearly what they're actually saying.

In short: a nice eye-catching little news piece, but no substance.

Cheers, Daniel.

[Daniel James]

Just checking.

A sensible precaution -- especially where disks are concerned. It would be nice it is was a sufficient precaution!

I'm not surprised to hear that of Maxtors, given their reputation ... but I am surprised at the actual numbers involved. I'd be interested to know how long your drives typically lasted before they failed.

Glad to hear it. Long may it remain so.

Cheers, Daniel.

[Geoff Beale]

Yes, and there are quite a few here who will also take every opportunity to tell us that Dyson vacs are crap too.

But there are doubtless many more folks who read these proclamations with smug smiles because their Maxtors and Dysons have given fault-free service for years. You don't hear from them though because they probably can't be bothered arguing with you.

[Andy Hall]

If you think about what you have said, it isn't logical.

I described a situation of a substantial number of Maxtor drives from different batches failing in a well controlled commercial environment.

Following that situation, it isn't very likely that I would buy their products for home use. OTOH, if I had bought one drive for home use and it had failed I would get a replacement. I might even do that a second time, but that would be it. The usage and analysis is quite different.

Equally, if 200 customers had individually bought these drives, there's a reasonable probability that 30 or so would have duds during the first year of operation. They might return them and by that time buy a diffierent product anyway. Thus for an individual, the issue might have gone unnoticed. It would depend also on the degree of pain that there would be in recovering the data onto the new drive. That affects people in different ways. In a commercial setting it was relatively expensive.

It's rather pointless mentioning vacuum cleaners, although some of the analogy applies. One can argue that failure of one vacuum cleaner is anecdotal and that comment from a commercial user is not that relevant for a domestic user whose use will be lighter anyway - i.e. it's likely that the domestic user will get a better experience than a commercial user attempting to use a domestic product commercially.

The disk drive example is the reverse. In domestic use the drive is likely to be power cycled and will most likely not receive a temperature controlled environment. This is worse than the commercial use case that I described where the environment was well controlled and drives run continuously and is likely to result in a higher failure rate.

If 30 customers out of 200 have a dud drive experience, it does mean that there are 170 who are happy for the moment. However, in the context of what the results should be and generally are in the industry, the figure is lousy.

Given that there are at least 4 other major brands, one can easily make much lower risk purchasing decisions.

I don't really care what you buy.

[Clint Sharp]

In message , Andy Hall writes

As a matter of interest, which Maxtor drives did you buy?