https://bugzilla.novell.com/show_bug.cgi?id=335505#c14
--- Comment #14 from Robert Davies 2007-11-15 19:13:52 MST ---
Kernel updated to SuSE kernel-default 2.6.22.12-0.1; 2x60GB Seagates which have
passed DFT's. I suspected my "intermittent sluggish performance, and machine
lock ups, after it had run a few hours" were caused by SuSE kernel 2.6.22.9-0.4
brown paper bag release.
So I ran torture test on disks bonnie++ & dbench, including simulataneous on
both disks on Host controller, tests found nothing system remained responsive
despite heavy CPU and saturated I/O usage; just 1 SMART Prefailure warning
about Raw_Read_Error_Rate (actually 62 to 63, then later 54 to 55), and some
Hardware ECC recovered messages. I save smartctl -a output, and will attach.
So swapped the newer disk sdb onto Promise controller - PDC20267
(FastTrak100/Ultra100) (rev 02)
When swap partition activated on the disk placed on Promise controller,
immediately ATA3 aka sdb caused Exceptions and the link was reset, without any
need to run benchmark, much logging caused, though a partition was mountable
from that disk. Error log attached previous comment.
On starting test bonnie++ mounting partion from disk on Promise controller, the
machine immediately seemed to freeze up, and there was a logged message, saying
"Clocksource tsc unstable (delta = 1405645341 ns)".
So this problem *IS* easily reproducible with latest SuSE kernel 2.6.22.12, you
just need the controller, and a disk on it. The problem has moved with the
controller not any particular hard drive.
The machine is practically unusable when Disk is accessed attached to the
Promise.
Please can I provide any info, to progress this Bug investigation, it seems
that it is impractical for th.sch@gmx.net to do so!
--
Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.