scsi hdd - taking a dump?

[bob4432]

recently i have noticed a significant slowdown with my system response. programs would take a lot more time to load, transfers would take longer and i noticed the system hdd (10k scsi) was working much more. virus/mal/adware tests have been done and report nothing. everything appears normal in the device manager. the machine is sluggish.

i ran a hdtach on it and it is indeed a lot slower than what is use to be - 44MB/s str, 11.3ms access and the graph spikes like mad, almost hitting the bottom in transfer. not too long ago hd tach graphed it at 60MB/s str, 8ms acces and the graph was gradual slowly going down with little spikes here and there.

would you say this is a card, cable or hdd issue? smart test says the drive is fine and it is defragged at least 1x weekly.

the 10K drive, at this point, is slower than my old 80GB pata......

 

[Lord Evermore]

Perhaps its developed bad sectors, and has remapped those sectors to the spare areas of the disk. Keeps your system from failing hard, but would probably result in performance loss if the head had to jump to a different place whenever it needed to read those sectors. The SCSI controller should have an option to verify the disk, but of course the system has to go down for that (unless the software for the controller allows it to be done live, I dunno).

SMART test? SCSI drives don't have SMART do they?

 

[bob4432]

Originally posted by: Lord Evermore
Perhaps its developed bad sectors, and has remapped those sectors to the spare areas of the disk. Keeps your system from failing hard, but would probably result in performance loss if the head had to jump to a different place whenever it needed to read those sectors. The SCSI controller should have an option to verify the disk, but of course the system has to go down for that (unless the software for the controller allows it to be done live, I dunno).

SMART test? SCSI drives don't have SMART do they?

as far as scsi and smart, not really sure. i know that i had a program, can't remember the name, but everybody uses instead of mbm, that pulled what i though was the smart info from the hdds to show errors and temps, etc.

i have seen the controller boot screen where you are talking about verify, but was not sure if it would destroy the data. can the bad sectors be fixed? i have a good image of the drive so i could format it and then put the image back on if you think that will help?? should this drive be warrantied or is this a normal process? i am pretty sure it still under warranty, but would definately verify.

 

[Lord Evermore]

Most drives use remapping to hide bad sectors by mapping them to spare areas of the disk. Here's some info.

Any bad sectors are a bad thing. New drives should have no bad sectors. If there are a few due to an unforeseen defect in the drive platter, that show up after a certain amount of active usage, but then no more develop, then it's usually fine. But if you have a few more creeping up every once in a while, then there is something causing them which will eventually probably result in total failure. If they've gotten so bad that they're affecting performance, you might be able to open the drive and see the grooves in the platter from a head dragging (hyperbolic but maybe possible).

I can't remember offhand whether the SCSI verify from the controller does a read/write/read type test which would wipe out your data. Check with the controller manufacturer/manual.

A chkdsk /r in Windows (technically on a reboot and outside the GUI presuming it's your boot disk) would scan the partition including all free space, and report any bad sectors, but may not be as thorough as the SCSI verify, or the tests using the drive maker's utilities if available.

Bad sectors of course is just a possible cause of the problems you're having, but it's the best I can think of.

 

[bob4432]

i just did a chksk /r and then ran hdtach - it is back up to normal str, cpu usage and access time....i just unloaded a sh!tload of software but did a defrag and thought that would take of it. would this mean the drive is health or should i order up a new one and send this in for rma?