Bug 30513
| Summary: | iostatistics not working with offboard PCI IDE controller (promise) | ||
|---|---|---|---|
| Product: | [Retired] Red Hat Linux | Reporter: | Ed McKenzie <eem12> |
| Component: | kernel | Assignee: | Arjan van de Ven <arjanv> |
| Status: | CLOSED RAWHIDE | QA Contact: | David Lawrence <dkl> |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 7.1 | ||
| Target Milestone: | --- | ||
| Target Release: | --- | ||
| Hardware: | i386 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2001-03-13 10:19:47 UTC | Type: | --- |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
|
Description
Ed McKenzie
2001-03-04 07:11:36 UTC
are you sure you are doing IO while running iostat? Works fine for me here. start iostat 1 in one window and then run say "find / -name foo" in another. I think you will see it works fine. Nope, it's still not displaying. Running find on both IDE drives shows up zero
in all counters except the CPU usage, which is about where it usually is for
this kind of activity.
Incidentally, I've also noticed that procinfo no longer shows disk block stats
in the right column, as it did in RH7.
'iostat 1' log attached. A single instance of find / was running for the entire
run.
[eem12@eem12 eem12]$ iostat 1
Linux 2.4.1-0.1.14 (eem12.resnet.cornell.edu) 03/05/2001
avg-cpu: %user %nice %sys %idle
29.36 13.92 4.09 52.63
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
dev3-0 0.00 0.00 0.00 18 0
dev11-0 0.00 0.00 0.00 32 0
avg-cpu: %user %nice %sys %idle
34.00 0.00 0.00 66.00
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
dev3-0 0.00 0.00 0.00 0 0
dev11-0 0.00 0.00 0.00 0 0
avg-cpu: %user %nice %sys %idle
33.00 0.00 0.00 67.00
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
dev3-0 0.00 0.00 0.00 0 0
dev11-0 0.00 0.00 0.00 0 0
avg-cpu: %user %nice %sys %idle
29.00 0.00 7.00 64.00
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
dev3-0 0.00 0.00 0.00 0 0
dev11-0 0.00 0.00 0.00 0 0
avg-cpu: %user %nice %sys %idle
33.00 0.00 1.00 66.00
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
dev3-0 0.00 0.00 0.00 0 0
dev11-0 0.00 0.00 0.00 0 0
avg-cpu: %user %nice %sys %idle
33.00 0.00 1.00 66.00
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
dev3-0 0.00 0.00 0.00 0 0
dev11-0 0.00 0.00 0.00 0 0
avg-cpu: %user %nice %sys %idle
35.00 0.00 1.00 64.00
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
dev3-0 0.00 0.00 0.00 0 0
dev11-0 0.00 0.00 0.00 0 0
avg-cpu: %user %nice %sys %idle
31.00 0.00 5.00 64.00
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
dev3-0 0.00 0.00 0.00 0 0
dev11-0 0.00 0.00 0.00 0 0
avg-cpu: %user %nice %sys %idle
34.00 0.00 1.00 65.00
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
dev3-0 0.00 0.00 0.00 0 0
dev11-0 0.00 0.00 0.00 0 0
avg-cpu: %user %nice %sys %idle
32.00 0.00 5.00 63.00
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
dev3-0 0.00 0.00 0.00 0 0
dev11-0 0.00 0.00 0.00 0 0
avg-cpu: %user %nice %sys %idle
27.00 1.00 2.00 70.00
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
dev3-0 0.00 0.00 0.00 0 0
dev11-0 0.00 0.00 0.00 0 0
[eem12@eem12 eem12]$
btw, this was on a freshly booted box, so the fs cache wasn't primed. Looking at the device numbers, it seems iostat is ignoring the PCI IDE board. dev11-0 is the scsi CD writer, and dev3-0 is the IDE CD-ROM. Both of those show up properly if IO occurs on the respective devices. Of course, iostat would be even more useful if it showed statistics for my hard disks. :) Are off-board IDE boards not going to be fully supported by system tools in RH7.1? I would hope they would, since several popular on-board chipsets are buggy and can be dangerous to use in many cases. iostat is manifesting the problem, but the real problem, if any, lies in the kernel. mentioning the device on the commandline (iostat -x /dev/hde) seems to work here, investigatin why it doesn't find them by default.... The kernel was indeed not accounting "high" IDE channels. Fixed for our next build. |