Summary: | app-admin/sysstat-8.0.3 + sys-kernel/gentoo-sources-2.6.24 : Pentium-M doesn't have 2 processors | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Toralf Förster <toralf> |
Component: | New packages | Assignee: | Jeroen Roovers (RETIRED) <jer> |
Status: | RESOLVED FIXED | ||
Severity: | minor | ||
Priority: | High | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Bug Depends on: | |||
Bug Blocks: | 207383 | ||
Attachments: |
kernel .config
content of /sys/devices/system/cpu/ |
Description
Toralf Förster
2008-01-28 13:27:56 UTC
Could you post your 2.6.24 kernel config as well, please? I am not at all sure this is a sysstat bug yet. Created attachment 142040 [details]
kernel .config
(In reply to comment #2) > Created an attachment (id=142040) [edit] > kernel .config Thanks. Nothing weird there. One thing though: did you clear out the old 8.0.2 generated files in /var/log/sa/? There may have been a format change that 8.0.3 didn't take into account. In other words, could you delete or move the old files and test again, please. I run /etc/init.d/sysstat at default level which do this. I contacted upstream already both for this bug (and BTW some more times in the past for some problems I had handling sysstat files got from AIX sysstems with kSar). The only thing I'm wondering is that these settings : n22 /var/log/sa # cat /etc/sysconfig/sysstat # How long to keep log files (in days). # If value is greater than 28, then log files are kept in # multiple directories, one for each month. HISTORY=7 # Compress (using gzip) sa and sar files older than (in days): COMPRESSAFTER=10 seems not be be recognozed b/c I had more than the specified 7 files : n22 /var/log/sa # l sa09 sa10 sa11 sa12 sa13 sa14 sa15 sa16 sa17 sa18 sa19 sa20 sa21 sa22 sa23 sa24 sa25 sa26 sa27 sa28 sa29 Got from upstream: Toralf Förster wrote: > > tfoerste@n22 ~ $ cat /proc/stat > cpu 39432 218600 5652 2271 3931 261 85 0 0 > cpu0 39432 218600 5652 2271 3931 261 85 0 0 > intr 3025952 2654713 3720 0 0 0 0 0 0 0 4770 0 226949 648 0 135097 55 > ctxt 1508631 > btime 1201538786 > processes 15564 > procs_running 3 > procs_blocked 0 > tfoerste@n22 ~ $ mpstat -P ALL > Linux 2.6.24-gentoo-fgcs (n22) 01/28/08 > > 21:33:23 CPU %user %nice %sys %iowait %irq %soft %steal %idle intr/s > 21:33:23 all 14.57 80.92 2.09 1.45 0.10 0.03 0.00 0.84 222.68 > 21:33:23 0 14.57 80.92 2.09 1.45 0.10 0.03 0.00 0.84 222.68 > 21:33:23 1 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 > tfoerste@n22 ~ $ sar -P ALL 0 > Linux 2.6.24-gentoo-fgcs (n22) 01/28/08 > > 21:33:27 CPU %user %nice %system %iowait %steal %idle > 21:33:27 all 14.55 80.94 2.22 1.45 0.00 0.84 > 21:33:27 0 14.55 80.94 2.22 1.45 0.00 0.84 > 21:33:27 1 0.00 0.00 0.00 0.00 0.00 0.00 > > IMO, if you look into your /sys/devices/system/cpu/ directory, you will see two subdirectories named cpu0 and cpu1, which means that your machine actually has two processors. But as you can notice, the lines printed by mpstat and sar for processor 1 display only null statistics. Sar and mpstat manual pages say that "On SMP machines a processor that does not have any activity at all (0.00 for every field) is a disabled (offline) processor." So to put it in a nutshell : your machine has 2 processors, and one of them has been set offline. Regards, -- Sébastien Godard (sysstat <at> orange.fr) Help support sysstat...! Visit http://pagesperso-orange.fr/sebastien.godard/ Created attachment 142182 [details]
content of /sys/devices/system/cpu/
(In reply to comment #6) > Created an attachment (id=142182) [edit] > content of /sys/devices/system/cpu/ Er, so are you saying that this bug is invalid or that Sébastien is wrong? Got this reply from upstream: >OK, I had never seen such a configuration ;-) >Please can you try this version of sysstat which should fix the bug? > >Thanks. and version 8.0.4.1, but can't test it today. tested upstream version 8.0.4.1 successfully :-) reopening for 8.0.4.1 inclusion in portage please also try and get this marked stable before 2.6.24 goes stable (In reply to comment #10) > reopening for 8.0.4.1 inclusion in portage > please also try and get this marked stable before 2.6.24 goes stable 1) This is a minor issue. 2) 8.0.4.1 was never published. |