Signs of a failing SCSI controller?

[Halz]

Controller is an Adaptec 29160, 3.10 firmware (most current), drives are IBM of mixed models.. DDYS, DNES, DMVS, (more, but don't remember the models off-hand)

Symptoms are that a drive (operating under normal interface speeds, and default controller settings, the only drive on the bus, or even among many drives) will intermittently emit an audible high-pitch sound for a short duration; 2 seconds or less..

For at least a year with this setup, there were no instances of drive failure and there was no audible pitch from the drives.. One of the drives eventually failed, suspectedly due to power-supply problem (Enermax 650Watt) giving an inadequate supply of power to the drives when they were under load.
About two years later now, five RMAed drives from IBM, I'm beginning to suspect the controller.. however, I haven't come across any utilities that could allow testing for a failing controller card, much less have experienced a controller failing other than outright failing; drives not being usable, at all.


Maybe one of the 9,699,481 members here has experienced a not-all-the-way up-to snuff SCSI controller and can relate some observations..

 

[Crusty]

123324 users are registered to the AnandTech forum.

There are currently 9,700,710 messages in the database.


Hehe, you got your numbers wrong. But one way to test your controller is to a) try it in another computer or b) try a different controller and see if you get the same problems.

 

[JC]

Hmm....I use one of those host adapters, it's never given me any trouble. I have used IBM DDYS drives with it, no noises, but both of them failed (one developed more and more bad sectors, the other just stopped reading/writing one day. I personally think it's just the IBM drives, I'll be using another brand in the future.

JC

 

[dkozloski]

I have been using SCSI drives exclusively for more than 10 years. I have found that Seagate are the best and the rest are varying grades of junk. With everyone else the company goes broke or disappears just about the time you need an RMA or support. At one time I had a stack of failed Quantum drives a foot high. Next Micropolis disappeared. Rodime, Marksman, Conner, CDC, Pertec, and a whole host of others have bit the dust but Seagate has withstood the test of time.

 

[Alptraum]

Originally posted by: dkozloski
I have been using SCSI drives exclusively for more than 10 years. I have found that Seagate are the best and the rest are varying grades of junk. With everyone else the company goes broke or disappears just about the time you need an RMA or support. At one time I had a stack of failed Quantum drives a foot high. Next Micropolis disappeared. Rodime, Marksman, Conner, CDC, Pertec, and a whole host of others have bit the dust but Seagate has withstood the test of time.

Oh my god. I haven't heard the name Micropolis in a long time. And I just found an 850meg Conner in my storage room, lol. I agree with you on Seagate SCSI quality. In the early days I thought Seagate sucked bad. One time out of a shipment of 1000 drives we stopped counting bad ones around 600 and just sent the whole shipment back (these were IDE though, back when in the IDE arena Seagate = crap and Conner = great. Lol, I remember when Seagate bought Conner my co-workers and I all almost cried). However, they have improved to the point that for the last 6 or 8 years I have used nothing but Seagate SCSI drives for both personal and work purchasing.

As far as your problem Halz, I would look at the drives before the controller. While I only deal with Seagate SCSI for the most part some of my clients and past jobs have had IBMs and in general they get a big thumbs down from me. I suppose its possible that the controller is involved, but out of all the bad controllers I have seen none of them have created symptoms like what you are describing.

You could try switching controllers but if this is an intermittent problem that way could take forever to figure out. I would suspect the drives before anything. Maybe contact IBM and see if they have ever heard of a controller causing what you are experiencing? I haven't heard of that at least.

 

[Halz]

I have the system running now with a borrowed (thanks JanSolo! <3 ) 19160 card running in place of the 29160 and the symptoms persist.. sooo, I'll just box these drives up and submit them for yet another RMA.. I'm at least grateful of IBM's 5 year warranties on the Ultrastar line of hard drives

..terrible that I had been so lazy as to not have tried a different SCSI card up until now. Meh