jaf@inference.com (Jose A. Fernandez) (04/29/91)
Folks, Has anyone experienced the following problem? Robert has informed us that there is a known problem that can cause internal disks in Sun workstations to die after being turned off! The multimedia/graphics group has been bitten twice by this in one week. The problem has to due with the internal lubricant in the drive drying out. When the disk is turned off, it can't be started again. The work-around (until the dirves can be replaced) is simple: 1) Don't turn off your Sun workstation! 2) If you MUST turn it off, be sure that you back up all critical work from your internal disks beforehand. If you've experienced the aforementioned problem, I'd appreciate hearing from you. It's always wise to have good reasons for nuking your local Sun office. Thanks, in advance, for your assistance.
hugh@slee01.srl.ford.com (Hugh Fader) (05/10/91)
jaf@inference.com (Jose A. Fernandez) writes: > Has anyone experienced the following problem? > > Robert has informed us that there is a known problem that can cause > internal disks in Sun workstations to die after being turned off! I can verify this. We have a four-day weekend for Easter for which we shutdown our machines -- SS1s with 2 100 MB Maxtor disks in each. When we came back one of our machines wouldn't boot due to disk failure. Fortunately, we have a maintenance contract. The Sun technician came out the same day and said he was swamped with service calls for this problem. Apparently the actuator sticks on some of the Maxtor disks when they are shut down for more than a short time. The technician's first attempt at a fix was to pull out the offending disk and bang on it with a telephone receiver. This had worked at his last three calls, he said. It didn't work for us however. He had to replace the disk. If you ask me, Sun should replace free of charge any disks exhibiting this problem -- maintenance contract or not. Hugh Fader hugh@slee01.srl.ford.com
beaulieu@uunet.uu.net (Larry Beaulieu) (05/10/91)
>Robert has informed us that there is a known problem that can cause >internal disks in Sun workstations to die after being turned off! The >multimedia/graphics group has been bitten twice by this in one week. > >The problem has to due with the internal lubricant in the drive drying >out. When the disk is turned off, it can't be started again. The We've had this problem with the 104MB internal hard drives. When repowering up the system the system waits for the drive to spin up and then barfs. The heads are apparently stuck in their home position and the system times out waiting for the drive to spin up. You can either: 1) Replace the drive or: 2) Take the drive out of the workstation and rap it once on the *side* against the edge of the desk. This will jar the heads loose and allow the drive to spin up again. Option 2 was relayed to me by one of the local Sun FS engineers as being a workaround solution given to him and his cohorts at the local office. I have done option 2) several times w/o any damage in order to get a system back on line quickly until the drive can be replaced but I wouldn't recommend the practice unless you were really stuck. Larry Beaulieu The usual disclaimers apply. beaulieu@gca.com
eggert@twinsun.com (Paul Eggert) (06/05/91)
The problem is called ``stiction'', it's well known in disk circles, and it's
not limited to Sun. Sometimes you can fix it by briskly rotating the drive by
hand about a quarter turn around a vertical axis, or by patting the drive once
gently (like a baby's bottom, I've been told). Why does this work? The rest
of this posting is quoted from Trudy E. Bell, ``Incredible shrinking
computers,'' _IEEE Spectrum_ 28, 5 (May 1991), 37-41; this well-written
article also covers the most likely solution to this problem in future drives,
namely dynamic loading. Bell writes:martin@knightsbridge.media.mit.edu (Martin Friedmann) (06/05/91)
jaf@inference.com (Jose A. Fernandez) writes: > Has anyone experienced the following problem? hugh@slee01.srl.ford.com (Hugh Fader) writes: > The technician's first attempt at a fix was to pull out the offending disk > and bang on it with a telephone receiver. Yes... I too will verify this... The 100 MEG maxtor drives can stick when powered down. We had one which wouldnt spin up... At first, I thought that the disk was a goner. We were planning on throwing it on the heap. So I opened the drive, breaking the seal, what the hell, and started fiddling with it. People around the lab started staring into it as it was the first time that we had seen the internals of a sun hard drive. People went as far as to fake "getting" dust particles into it and fake spitting into it. But miraculously I booted the sparcstation while holding the open disk drive in my hand watching the heads seek back and forth. I "resealed" the drive and put the machine back together and it all still works to this day!!! This happened to another of our drives, making the stats 2 drives frozen for a total of about 3 instances of frozen drive in one year of "normal" power cycles for 6 machines with about 10 drives total... Marty.
hugh@slee01.srl.ford.com (Hugh Fader) (06/05/91)
hugh@slee01.srl.ford.com (Hugh Fader) writes: > > I can verify this. We have a four-day weekend for Easter for which we > shutdown our machines -- SS1s with 2 100 MB Maxtor disks in each. When we An alert correspondent pointed out that I was mistaken. These are not Maxtor disks, but Quantum disks. He also wrote that slowly warming the disk up rather than whacking it might fix the stuck actuator problem. Sorry about any confusion I may have caused. -- Hugh Fader hugh@slee01.srl.ford.com
senese@schug.larc.nasa.gov (Fred Senese) (06/28/91)
Larry Beaulieu (gca!beaulieu@uunet.uu.net) writes: > When repowering up the system the system waits for the drive to spin up > and then barfs. The heads are apparently stuck in their home position and > the system times out waiting for the drive to spin up. > You can either > 1) Replace the drive > or: > 2) Take the drive out of the workstation and rap it once on the *side** > against the edge of the desk. This will jar the heads loose and > allow the drive to spin up again. This problem was common with Seagate SCSI drives a few years ago. It was attributed to ``worms'' of lubricant forming on the disk surface after several months of spinning; the extra lubricant forms a seal and locks the heads in home position. I replaced one such drive and the replacement began to display the same behaviour after about six months. I've had it for about two years now. When it fails to spin up, I carefully nudge the spindle with a screwdriver. The drive always spins up after a single nudge, and with my setup I don't have to pull the drive off its mounting rails. --