Hello, I've noticed that top now is now unreliable in measuring system performance. [timon_at_timon timon]$ uname -mv FreeBSD 5.2-CURRENT #19: Sat Mar 13 16:27:48 MSK 2004 root_at_timon.nist:/usr/obj/usr/src/sys/TIMON i386 options SCHED_ULE (cvsupped just before rebuilding kernel & world) top gives weird CPU/WCPU usage: only first (rarely, first couple) of processes displayed have non-zero CPU/WCPU usage, all others have zero cpu usage, but totals, shown in second line of top output differs from sum of values in CPU column alot. Same with ps. E.g.: 51 processes: 2 running, 49 sleeping CPU: 20.2% user, 0.0% nice, 3.5% system, 2.3% interrupt, 74.0% idle Mem: 168M Active, 141M Inact, 98M Wired, 716K Cache, 60M Buf, 91M Free Swap: 231M Total, 231M Free PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU CMD 1363 timon 77 0 49544K 42140K RUN 1:11 6.25% 6.25% mozill 890 timon 77 0 100M 83152K select 1:21 3.12% 3.12% XFree8 1344 timon 20 0 16184K 13144K kserel 1:55 0.00% 0.00% xmms 918 timon 20 0 78616K 50344K kserel 0:44 0.00% 0.00% evolut 901 timon 76 0 5556K 2604K select 0:25 0.00% 0.00% wmusic 919 timon 20 0 41304K 37836K kserel 0:21 0.00% 0.00% pan 900 timon 8 0 2196K 1616K nanslp 0:06 0.00% 0.00% wmWeat 899 timon 8 0 2196K 1616K nanslp 0:06 0.00% 0.00% wmWeat 895 timon 76 0 6552K 5872K select 0:05 0.00% 0.00% wmakerReceived on Sun Mar 14 2004 - 06:09:38 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:47 UTC