Hello, Thanks for your answer. The chipset is an Intel ICH4. Unfortunetaly I didn't find similar problems with that configuration. I was wondering whether it would be more stable to run with DMA deactivated ? Performances would suffer but I'm much more concerned about stability than performance. Is is safe to run "hdparm -d0 /dev/hda" on a live system ? Thanks for your help, Tom -----Original Message----- From: Willem van der Walt<willem@xxxxxxxxxxxxxxxxx> [mailto:willem@xxxxxxxxxxxxxxxxx] Sent: lundi 10 novembre 2003 10:36 To: 'redhat-list@xxxxxxxxxx' Subject: Re: Input/output error lerly as you say, either a bad disk, controler or an unsupported hardware issue. I have seen simmelar messages on a hard drive that was going. That one started to require more and more fsck runs before it died. The machine also became slower and slower as time went on. With a service provider like that, you should most likely have to find there problem for them. When you can, check dmesg to see if you can see what it says about the IDE controler and check on google. If you search on the chip number and linux and error you should get something. Upgrading the kernel might help too. Regards, Willem On Mon, 10 Nov 2003, Leseney Thomas wrote: > Hello, > > I had the following problems a few months ago: the server I rent gets stuck every now and then. > Running any command that requires write access to the disk results in > "Input/output error". The server is running RedHat 8.0 (kernel 2.4.20). > > Some of them asked for logs but unfortunately I had none at this point since the server > could not write anything :-( > > This problem happened again a few days ago and, this time, I found some relevant logs: > > in dmesg: > EXT3-fs error (device ide0(3,3)) in ext3_reserve_inode_write: IO failure > end_request: I/O error, dev 03:03 (hda), sector 0 > end_request: I/O error, dev 03:03 (hda), sector 264056 > > in var/log/messages: > > Oct 24 19:55:26 ws33 kernel: hda: dma_timer_expiry: dma status == 0x21 > Oct 24 19:55:43 ws33 kernel: hda: timeout waiting for DMA > Oct 24 19:55:43 ws33 kernel: hda: timeout waiting for DMA > Oct 24 19:55:43 ws33 kernel: hda: (__ide_dma_test_irq) called while not waiting > Oct 24 19:55:43 ws33 kernel: hda: status error: status=0x58 { DriveReady SeekComplete DataReque > st } > Oct 24 19:55:43 ws33 kernel: > Oct 24 19:55:43 ws33 kernel: hda: drive not ready for command > Oct 24 20:38:18 ws33 kernel: hda: irq timeout: status=0xd0 { Busy } > Oct 24 20:38:18 ws33 kernel: > Oct 24 20:38:53 ws33 kernel: ide0: reset timed-out, status=0x80 > Oct 24 20:38:53 ws33 kernel: hda: status timeout: status=0x80 { Busy } > Oct 24 20:38:53 ws33 kernel: > Oct 24 20:38:53 ws33 kernel: hda: drive not ready for command > Oct 24 20:39:22 ws33 kernel: ide0: reset: success > Oct 24 20:39:56 ws33 kernel: hda: irq timeout: status=0xd0 { Busy } > Oct 24 20:39:56 ws33 kernel: > Oct 24 20:40:29 ws33 kernel: ide0: reset timed-out, status=0x80 > Oct 24 23:01:39 ws33 kernel: hda: irq timeout: status=0xd0 { Busy } > Oct 24 23:01:39 ws33 kernel: > Oct 24 23:02:09 ws33 syslogd: /var/log/secure: Input/output error > Oct 24 23:02:09 ws33 kernel: ide0: reset timed-out, status=0x80 > Oct 24 23:02:09 ws33 kernel: end_request: I/O error, dev 03:03 (hda), sector 34255 > Oct 24 23:02:09 ws33 kernel: end_request: I/O error, dev 03:03 (hda), sector 34255 > Oct 24 23:02:09 ws33 kernel: end_request: I/O error, dev 03:03 (hda), sector 264336 > Oct 24 23:02:09 ws33 kernel: end_request: I/O error, dev 03:03 (hda), sector 264344 > Oct 24 23:02:09 ws33 kernel: end_request: I/O error, dev 03:03 (hda), sector 1056816 > Oct 24 23:02:09 ws33 kernel: end_request: I/O error, dev 03:03 (hda), sector 1180872 > Oct 24 23:02:09 ws33 kernel: end_request: I/O error, dev 03:03 (hda), sector 1229928 > Oct 24 23:02:09 ws33 kernel: end_request: I/O error, dev 03:03 (hda), sector 5242960 > Oct 24 23:02:09 ws33 kernel: end_request: I/O error, dev 03:03 (hda), sector 6029360 > [...] > > I am not familiar with hardware related failures but this looks like a hard disk failure or a driver > problem. My hosting provider still claims that it is due to the applications (apache ...) I installed. > > Any thoughts ? > > Thanks for your help, > > Tom > > -- redhat-list mailing list unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe https://www.redhat.com/mailman/listinfo/redhat-list -- redhat-list mailing list unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe https://www.redhat.com/mailman/listinfo/redhat-list
