[cobalt-users] Problems after Kernel upgrade on RAQXTR

["Mark Hudson [mch22]" <m.c.hudson@xxxxxxxxxx>]

Ever since installing the kernel upgrade
RaQXTR-All-Kernel-1.0.1-2.2.16C35_V-1.pkg on our two RAQXTRs they are
both exhibiting RAID & IDE problems...


Apr 12 04:09:44 ltsraq1 kernel: hde: write_intr error1: nr_sectors=8,
stat=0xd0
Apr 12 04:09:44 ltsraq1 kernel: hde: write_intr: status=0xd0 { Busy }
Apr 12 04:09:44 ltsraq1 kernel: ide2: reset: success
Apr 12 04:09:44 ltsraq1 kernel: hde: write_intr error1: nr_sectors=8,
stat=0xd0
Apr 12 04:09:44 ltsraq1 kernel: hde: write_intr: status=0xd0 { Busy }
Apr 12 04:09:44 ltsraq1 kernel: ide2: reset: success
etc....

Then it re-boots itself and you get this:

Apr 12 06:28:48 ltsraq1 kernel: hdi: write_intr error1: nr_sectors=8,
stat=0xd0
Apr 12 06:28:48 ltsraq1 kernel: hdi: write_intr: status=0xd0 { Busy }
Apr 12 06:28:48 ltsraq1 kernel: ide4: reset: success
Apr 12 06:28:48 ltsraq1 kernel: hdi: write_intr error1: nr_sectors=8,
stat=0xd0
Apr 12 06:28:48 ltsraq1 kernel: hdi: write_intr: status=0xd0 { Busy }
Apr 12 06:28:48 ltsraq1 kernel: ide4: reset: success
etc... utill....
Apr 12 06:51:30 ltsraq1 kernel: hdi: timeout waiting for DMA
Apr 12 06:51:30 ltsraq1 kernel: hdi: 0 bytes in FIFO
Apr 12 06:51:30 ltsraq1 kernel: ide_dmaproc: chipset supported
ide_dma_timeout func only: 14
Apr 12 06:51:30 ltsraq1 kernel: hdi: irq timeout: status=0x50 {
DriveReady SeekComplete }

Then it stops responding completely so you have to power it off
uncleanly and get this:

Apr 12 08:18:32 ltsraq1 kernel: hdi: status timeout: status=0x80 {
Busy }
Apr 12 08:18:32 ltsraq1 kernel: hdi: DMA disabled
Apr 12 08:18:32 ltsraq1 kernel: hdi: drive not ready for command
Apr 12 08:18:32 ltsraq1 kernel: ide4: reset: success
Apr 12 08:18:32 ltsraq1 kernel: hdk: 0 bytes in FIFO
Apr 12 08:18:32 ltsraq1 kernel: ide_dmaproc: chipset supported
ide_dma_timeout func only: 14
Apr 12 08:18:32 ltsraq1 kernel: hdk: timeout waiting for DMA
Apr 12 08:18:32 ltsraq1 kernel: hdk: 0 bytes in FIFO

Then re-boot again and get this:

Apr 12 11:09:15 ltsraq1 kernel: hdk: write_intr error1: nr_sectors=1,
stat=0xd0
Apr 12 11:09:15 ltsraq1 kernel: hdk: write_intr: status=0xd0 { Busy }
Apr 12 11:09:15 ltsraq1 kernel: ide5: reset: success
Apr 12 11:09:48 ltsraq1 kernel: hdk: write_intr error1: nr_sectors=107,
stat=0xd0
Apr 12 11:09:48 ltsraq1 kernel: hdk: write_intr: status=0xd0 { Busy }
Apr 12 11:09:48 ltsraq1 kernel: ide5: reset: success

Continuously...... Errors on all disks....

This only started since we did the kernel upgrade. It happens on both
machines. We even replaced the hdg disk when it first happened, assuming
a faulty disk. We then swapped all the disks around between the two
XTRs, giving each the "personality" of the other. The problems remained.
And although on sun's description of the patch they say



"This patch will update the kernel to version 2.2.16C35. This kernel
addresses a RAID issue on the RaQXTR where synchronization of a RAID
array could take a long time"



It actually sometimes still takes a very long time (thousands of
minutes).



Anyone else seeing problems?



Mark