Acceptable Hard Disk Error Rates From Testing

[Peahippo]

(First posted on comp.hardware, but I received no response.)

I have a pile of IDE hard disks (1 to 4GB) at work that I need to run through some reliability testing to make reasonably sure they won't die or show intermittent problems while running. These hdds will go into computers primarily used for front-office banking, hence I am very concerned about reliability. (Added to this are the embarrassment and waste factors, in showing up at a banking center with a replacement that either dies immediately or very soon.)

I downloaded COREHDT for the testing, and it ran read/write tests on one particular 1.08GB drive. The test showed 65 errors noted in 1e+12 bits. (The errors were all in the "Read-MultiSector" and "Read-64KB" test modes.)

1. How many bits of testing are the minimum for measuring reliability in this way? I recall reading specs that commonly give error rates in the "x per 1e+12" (1 trillion bits) range. For running COREHDT on a 486, it takes about 24hr to run through 1e+12 bits.

2. How much error is tolerable? For instance, 4 errors per 1e+12 bits?

I've tested half a dozen drives this way, and others have 0 errors.

 

[PeterGriff]

I havent used this program but I would say that ANY errors should be a concern - especially if most of the other drives come up clean. I have always just used Norton Disk Doctor and done a thorough surface scan to look for errors. Your program would seem to be even more thorough by writing and reading back repeatedly. Any error would lead to data corruption which cant be good!. I presume that RAM errors wouldnt interfere with this test (are they are being tested in the same system, or in-situ in their own case?).

Peter G

 

[wpfaff]

Is there 65 errors or 65 bytes. Does it tell how many bits are affected? The problem is that 1.08 Gigs isn't very much capacity if this is a windoze machine. When you format this drive the formatter is going to mark all of the bad blocks and your going to have even less capacity. Put this harddrive up for sale on ebay to the highest bidder.

Your reputation and business goodwill is to important to lose on this transaction.

 

[tbirdkid]

I use Spinrite to check all the older drives that I encounter. Spinrite has the capability to re-low level format a drive as well. And after all that if there is ANY errors........hello E-Bay!

 

[Peahippo]

I'll address all the issues here so far.

Any errors are a concern, but with recent events all of us can understand, budgets are at least tight and I must make the most use of my resourcces. If the drive will run, then I will attempt to use it.

The hdds are being tested on one system, a standalone 486. If the 486 is causing the problems, then perhaps it would cause them on at least some of the disks I'm testing, and so far that's not the case.

I don't know how many bits were affected. The 65 errors occured as 64 in these modes: Read-MultiSector, Read-64KB. There was only 1 remaining error, for Read-SingleSector. It could have been that this particular drive does poorly in those modes, and it could also be that common OS usage of the hdd doesn't use those modes either.

Drive size is a non issue. We use drives that small for bank teller machines. They are minimal, although use Win95. We obsolete drives that are less than 1GB.

eBay is a great site, but I've no facility for using it for the job, since I can't get reimbursed for my expenses. Now, if there was a critical need, I'm sure the VP in my chain of command would sign off on an expense entry if I bought from eBay and put the equipment into service, but that's a rare beastie.

My main concern was about the reliability of the drive. We have teller machines crash "all the time" (not excessively, but often enough to notice), and most recover by being reloaded with the OS builds that we have.

Any thoughts?

 

[edfair]

Earlier machines would have enough errors that up to 10% of the sectors would get locked out, and nobody much got upset.

You probably need to check the manufacturer's websites for specs, I think I've seen the read recovery as one of the parameters, but that may also have been from long ago.

I would consider successive tests a better indicator. One before you install it, then one the next time the drive is replaced. Any mojor change is probably due to degredation of read electronics and probably going to get worse.



Ed Fair
Any advice I give is my best judgement based on my interpretation of the facts you supply. Help increase my knowledge by providing some feedback, good or bad, on any advice I have given.

 

[tbirdkid]

Well, I would still grab a copy of Spinrite and try refreshing those older platters to have a better attempt at reliablilty.

 

[tsurikov]

Just wondering... a few people above suggested Ebay for drives that have errors on them... now why would anyone want to buy a defective drive? (Or is the idea to lie about the defects... which is in itself not quite right..?)

--Michael

 

[edfair]

Well , at least one seller was honest, he advertized his products as "guaranteed dead hard drives".

I would suspect that probably 10% of the computer stuff sold on ebay is either defective or questionable. At the prices stuff goes for very few would ever complain.

Ed Fair
Any advice I give is my best judgement based on my interpretation of the facts you supply. Help increase my knowledge by providing some feedback, good or bad, on any advice I have given.