Post by thread:[OT:] hardware or software problem?

I'm having trouble on a computer, every few minutes certain things stop for 5 seconds. Specifically, anything that needs the hard drive has to wait. /var/log/messages shows lines that correspond to these events, like this: Dec 11 14:59:23 ctc kernel: hda: lost interrupt Each time it does this thing, it puts a line like that in the log. The machine is a fairly conventional old Pentium 2, 300MHz, 192MB ram, Tyan motherboard, 8MB ATI AGP video, SB16 ISA audio, tulip PCI network card, 40GB Maxtor hard drive. CDROM drive is on the secondary cable. However, it's fairly new to me, and has done it since I put it in use. It's running kernel 2.4.20. That's Linux, by the way. I don't really have a way to test it with Windows. This problem sometimes happens twice in a minute, sometimes only once in 15 minutes. The hard drive doesn't make any bad sounds or act bad in any way. Any suggestions where I should look? I am thinking the first thing to do is play around with PCI settings in the bios. Thanks, Bob -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email spam_OUTlistservTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body

Bob, I'm also having hard drive problems at home at the moment - but I am getting noises as well as delays. However, you can download harddrive analysers off the web. They use the SMART monitoring built into the hard drive which claims to be able to identifty 70% of all drive problems before they become fatal. You may need to enable the SMART monitor from your bios/setup. Most of the free downloadables are limited time (30 day) but diskcheckup at http://www.passmark.com is free for home users. (Costs are pretty low for full versions anyway). Interpreting the results is a bit more difficult however but it should point you in the right direction. Apparently a common source of problems is the IDE connectors coming loose over time - might be worthwhile checking? Or you can download free diagnostic software from the disk manufacturer e.g seatools from http://www.seagate.com. This seems to work but only told me I had a failure flag set &that I needed to backup my drive & replace it ASAP. (wonderful - just 2 months outside the 1 year warranty !!). The long diagnostic check I left to run overnight & it froze with no results at all. Richard P I'm having trouble on a computer, every few minutes certain things stop for 5 seconds. Specifically, anything that needs the hard drive has to wait. /var/log/messages shows lines that correspond to these events, like this: Dec 11 14:59:23 ctc kernel: hda: lost interrupt Each time it does this thing, it puts a line like that in the log. The machine is a fairly conventional old Pentium 2, 300MHz, 192MB ram, Tyan motherboard, 8MB ATI AGP video, SB16 ISA audio, tulip PCI network card, 40GB Maxtor hard drive. CDROM drive is on the secondary cable. However, it's fairly new to me, and has done it since I put it in use. It's running kernel 2.4.20. That's Linux, by the way. I don't really have a way to test it with Windows. This problem sometimes happens twice in a minute, sometimes only once in 15 minutes. The hard drive doesn't make any bad sounds or act bad in any way. Any suggestions where I should look? I am thinking the first thing to do is play around with PCI settings in the bios. Thanks, Bob -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email .....listservKILLspam@spam@mitvma.mit.edu with SET PICList DIGEST in the body -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email listservKILLspammitvma.mit.edu with SET PICList DIGEST in the body

Bob Blick wrote: > I'm having trouble on a computer, every few minutes certain things stop > for 5 seconds. Specifically, anything that needs the hard drive has to > wait. > > /var/log/messages shows lines that correspond to these events, like this: > > Dec 11 14:59:23 ctc kernel: hda: lost interrupt Bob, I'd try pulling the network card. See if it continues to do it without the Tulip card on the PCI bus. They're kinda notorious for being cheap, but 99.9% of the time they do "just work". The drive could be failing. You could also test hammering the disk I/O with something like hdparm speed tests and/or bonnie and see if it gets worse or other error messages pop up to give other clues. Another good check would probably be to boot to single user mode and force an fsck if you're using an ext2/3 filesystem on all your partitions. The fact that it's so variable in time suggests that something else on the bus is messing around intermittantly. If you have spare IDE cables lying around it might be worth swapping them out just to see. Reseating everything probably wouldn't hurt either... I'd do some Googling with your specific drive part number and motherboard type to see if anyone else experiences the problem. Also, 2.4.20 is pretty "new" but there's been a couple of revs of 2.4 after that... might be worth slapping a newer kernel on if the box is using a package manager that makes it simple to do. Good luck, Nate, .....nateKILLspam.....natetech.com p.s. A quick google search found a guy who's had a machine doing this for 4 years who can't find it, but the warnings he's received about the drive failing have never come true. If you were getting another error like this guy: lists.slug.org.au/archives/slug/2002/04/msg00313.html I'd say maybe the disk is starting to fail... but not always... -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email EraseMElistservspam_OUTTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body

Richard.Prosser@POWERWARE.COM wrote: > you can download harddrive analysers off the web. They use the SMART > monitoring built into the hard drive which claims to be able to identifty > 70% of all drive problems before they become fatal. You may need to enable > the SMART monitor from your bios/setup. Hey Bob, if you do want to do SMART tests and the disk supports it, there was an article on doing this in either Linux Format or Linux Journal this month. Neither are here at the office, but I can check which it was and post links from them if that'd be helpful. Nate, natespam_OUTnatetech.com -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email @spam@listservKILLspammitvma.mit.edu with SET PICList DIGEST in the body

Nate Duehr said: > I'd try pulling the network card. See if it continues to do it without > the Tulip card on the PCI bus. They're kinda notorious for being cheap, > but 99.9% of the time they do "just work". > > The drive could be failing. You could also test hammering the disk I/O > with something like hdparm speed tests and/or bonnie and see if it gets > worse or other error messages pop up to give other clues. > > Another good check would probably be to boot to single user mode and > force an fsck if you're using an ext2/3 filesystem on all your > partitions. > > The fact that it's so variable in time suggests that something else on > the bus is messing around intermittantly. If you have spare IDE cables > lying around it might be worth swapping them out just to see. Reseating > everything probably wouldn't hurt either... Good tips all of them, thanks. I looked back to the head of messages, and I have been getting them since during the linux install three weeks ago. They don't seem to be related to temperature, the messages come when the machine is first booted. Generally, the more disk activity, the more frequent. It wouldn't surprise me if the hard drive was at fault, but I'm used to seeing multiple types of messages when that happens, and all I get is that one, so that makes me wonder. Plus it doesn't seem to get worse or better with temperature. BTW the drive runs cool. Tonight I'm going to: Replace the cables, see what happens. Pull the network card, see what happens. If it still acts up, I'll put the hard drive and video card into my other computer, which has a different motherboard, and see what happens. At least I don't get greasy working on my computer :-) Thanks again for the help, I'll post my results. Cheerful regards, Bob -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email KILLspamlistservKILLspammitvma.mit.edu with SET PICList DIGEST in the body

----- Original Message ----- From: Bob Blick <RemoveMEbobblickTakeThisOuTCOVAD.NET> To: <spamBeGonePICLISTspamBeGoneMITVMA.MIT.EDU> Sent: Thursday, December 11, 2003 11:26 PM Subject: [OT:] hardware or software problem? {Quote hidden}> I'm having trouble on a computer, every few minutes certain things stop > for 5 seconds. Specifically, anything that needs the hard drive has to > wait. > > /var/log/messages shows lines that correspond to these events, like this: > > Dec 11 14:59:23 ctc kernel: hda: lost interrupt > > Each time it does this thing, it puts a line like that in the log. The > machine is a fairly conventional old Pentium 2, 300MHz, 192MB ram, Tyan > motherboard, 8MB ATI AGP video, SB16 ISA audio, tulip PCI network card, > 40GB Maxtor hard drive. CDROM drive is on the secondary cable. > > However, it's fairly new to me, and has done it since I put it in use. > It's running kernel 2.4.20. That's Linux, by the way. I don't really have > a way to test it with Windows. > > This problem sometimes happens twice in a minute, sometimes only once in > 15 minutes. The hard drive doesn't make any bad sounds or act bad in any > way. > > Any suggestions where I should look? I am thinking the first thing to do > is play around with PCI settings in the bios. > > Thanks, > > Bob Hi Bob, Identify the IDE chipset on your motherboard. Search through the Linux kernel docs for known problems with that chipset. Recompile the kernel to compensate for the bugs. Several years ago I discovered that the motherboard in my build server had a dodgy IDE chipset (I think it was an RZ1000 - or something like that). Since the machine used a separate SCSI subsystem and the IDE was never used, I did not learn about this IDE problem for years after the machine had been comissioned. If you look into building your own tailored kernel you will find config options to build in problematic chipset workarounds. I believe these are disabled by default since they have performance penalties and it is assumed the chipsets are rare these days. As a mater of interest, what distro are you using? Something else to consider looking at is "power save mode" (possibly APM). Is something in your bios or linux kernel trying to shutdown your HD? Regards Sergio Masci http://www.xcprod.com/titan/XCSB - optimising structured PIC BASIC compiler -- http://www.piclist.com hint: To leave the PICList TakeThisOuTpiclist-unsubscribe-requestEraseMEspam_OUTmitvma.mit.edu