linux server cpu idle time below 10%

oracle 8.1.7 and advance server linux

i just want to know how the cpu idle time is below 10%

when I execute a sar command i get this output
03:40:00 PM CPU %user %nice %system %idle
03:50:00 PM all 91.01 0.04 7.99 0.96
04:00:00 PM all 91.92 0.04 7.36 0.69
04:10:00 PM all 90.49 0.04 8.68 0.80
04:20:00 PM all 92.31 0.03 6.83 0.82
04:30:00 PM all 91.01 0.03 8.85 0.11
04:40:00 PM all 91.36 0.04 8.51 0.08
04:50:00 PM all 92.08 0.04 7.81 0.06
05:00:00 PM all 91.70 0.03 8.22 0.05
05:10:00 PM all 90.59 0.05 9.27 0.10
05:20:00 PM all 91.63 0.04 8.28 0.06
05:30:00 PM all 90.09 0.04 9.80 0.07
05:40:00 PM all 89.71 0.03 10.17 0.09
05:50:01 PM all 89.61 0.04 10.23 0.12
06:00:01 PM all 89.71 0.03 9.65 0.61
06:10:00 PM all 90.55 0.04 8.82 0.59
06:20:00 PM all 89.40 0.03 9.94 0.62
06:30:00 PM all 89.36 0.04 9.76 0.84
06:40:00 PM all 88.86 0.04 10.44 0.66
06:50:01 PM all 90.13 0.04 9.27 0.56
07:00:00 PM all 89.11 0.04 10.23 0.63
07:10:00 PM all 89.14 0.04 10.21 0.61
07:20:00 PM all 89.94 0.03 9.44 0.58
07:30:01 PM all 89.43 0.04 9.97 0.55
07:40:00 PM all 90.33 0.05 8.98 0.64
07:50:00 PM all 89.91 0.04 9.31 0.74
08:00:00 PM all 90.41 0.05 8.82 0.72
08:10:00 PM all 90.09 0.04 9.33 0.55
08:20:00 PM all 87.84 0.03 11.17 0.95

well i looked at the backend , file system free space is there, tablespace free size is there , free memmory is also there what could be the issue ?

[oracle@ORACLEAS bdump]$ free -m
total used free shared buffers cached
Mem: 3851 3836 14 1006 193 1448
-/+ buffers/cache: 2194 1657
Swap: 2047 622 1425

[oracle@ORACLEAS bdump]$ df -h
Filesystem Size Used Avail Use% Mounted on
/dev/cciss/c0d0p5 4.8G 884M 3.6G 19% /
/dev/cciss/c0d0p1 116M 14M 96M 13% /boot
/dev/cciss/c0d0p3 43G 25G 16G 59% /home1
/dev/cciss/c0d0p2 44G 4.5G 37G 11% /home2
none 1.9G 0 1.8G 0% /dev/shm
/dev/cciss/c0d0p9 482M 8.1M 449M 2% /tmp
/dev/cciss/c0d0p6 3.8G 1.1G 2.5G 30% /usr
/dev/cciss/c0d0p8 1.4G 93M 1.2G 7% /var

Re: linux server cpu idle time below 10%

Quote:

Originally posted by gajananpkini
oracle 8.1.7 and advance server linux

i just want to know how the cpu idle time is below 10%

when I execute a sar command i get this output
03:40:00 PM CPU %user %nice %system %idle
03:50:00 PM all 91.01 0.04 7.99 0.96
04:00:00 PM all 91.92 0.04 7.36 0.69
04:10:00 PM all 90.49 0.04 8.68 0.80
04:20:00 PM all 92.31 0.03 6.83 0.82
04:30:00 PM all 91.01 0.03 8.85 0.11
04:40:00 PM all 91.36 0.04 8.51 0.08
04:50:00 PM all 92.08 0.04 7.81 0.06
05:00:00 PM all 91.70 0.03 8.22 0.05
05:10:00 PM all 90.59 0.05 9.27 0.10
05:20:00 PM all 91.63 0.04 8.28 0.06
05:30:00 PM all 90.09 0.04 9.80 0.07
05:40:00 PM all 89.71 0.03 10.17 0.09
05:50:01 PM all 89.61 0.04 10.23 0.12
06:00:01 PM all 89.71 0.03 9.65 0.61
06:10:00 PM all 90.55 0.04 8.82 0.59
06:20:00 PM all 89.40 0.03 9.94 0.62
06:30:00 PM all 89.36 0.04 9.76 0.84
06:40:00 PM all 88.86 0.04 10.44 0.66
06:50:01 PM all 90.13 0.04 9.27 0.56
07:00:00 PM all 89.11 0.04 10.23 0.63
07:10:00 PM all 89.14 0.04 10.21 0.61
07:20:00 PM all 89.94 0.03 9.44 0.58
07:30:01 PM all 89.43 0.04 9.97 0.55
07:40:00 PM all 90.33 0.05 8.98 0.64
07:50:00 PM all 89.91 0.04 9.31 0.74
08:00:00 PM all 90.41 0.05 8.82 0.72
08:10:00 PM all 90.09 0.04 9.33 0.55
08:20:00 PM all 87.84 0.03 11.17 0.95

well i looked at the backend , file system free space is there, tablespace free size is there , free memmory is also there what could be the issue ?

[oracle@ORACLEAS bdump]$ free -m
total used free shared buffers cached
Mem: 3851 3836 14 1006 193 1448
-/+ buffers/cache: 2194 1657
Swap: 2047 622 1425

[oracle@ORACLEAS bdump]$ df -h
Filesystem Size Used Avail Use% Mounted on
/dev/cciss/c0d0p5 4.8G 884M 3.6G 19% /
/dev/cciss/c0d0p1 116M 14M 96M 13% /boot
/dev/cciss/c0d0p3 43G 25G 16G 59% /home1
/dev/cciss/c0d0p2 44G 4.5G 37G 11% /home2
none 1.9G 0 1.8G 0% /dev/shm
/dev/cciss/c0d0p9 482M 8.1M 449M 2% /tmp
/dev/cciss/c0d0p6 3.8G 1.1G 2.5G 30% /usr
/dev/cciss/c0d0p8 1.4G 93M 1.2G 7% /var

whats tablespace size and free space got to do with cpu usage? All you cpu's are being used - whats the problem?

i dont have much experuience of linux admin so asked this.

so should a cpu idle % be less or it should be higher always

i mean between 60 to 70% or less than 10%

100% CPU Usage on a tuned database is nothing to worry about. It just means you are using all available CPU to get your work done.

That being said, there are very few cases where all the SQL on an instance is tuned. My guess is you have some bad queries sucking a lot of CPU that need to be tuned or you have a really large user population. I would look at your run queue to determine if there are jobs waiting.

check iostat. cpu may be waiting on io