CPU avg load

[Hector Vargas]

Hello forum,


I have a hp server 4340 with 4 CPU and it is most of the time running on ~7 cpu load and my client claims that number is to high. My question is what is consider an acceptable CPU average load for a server with 4 CPUs and if there is any documentation out there that I can use to back it up.
thanks,

[Paul Sperry]

A load of 7 does seem high. I have a system with an oracle database that thousands of people connect to it's usually at 0.7 to 1.0
as far as load goes. What does top show you?

[Geoff Wild]

Well - that is high - depending on what the server is doing...ideal is less then 1.

How much ram is in it?

There are a ton of performance checks you can do.

Things like kernel parms - dbc_max_pct...

Maybe the system is swapping? IE not enough ram...

Do you have glance and measureware?


I run a report like so daily from cron:

/opt/perf/bin/extract -xp -r /home/gwild/zgbltemplate -g -b today 7:00 -e today 18:00 -f stdout | /bin/mailx -s 'system performance report' sysadm >/dev/null 2>&1


# cat /home/gwild/zgbltemplate
REPORT "MWA Export on !SYSTEM_ID"
FORMAT ASCII
HEADINGS ON
SEPARATOR="|"
SUMMARY=60
MISSING=0
DATA TYPE GLOBAL
YEAR
DATE
TIME
GBL_CPU_TOTAL_UTIL
GBL_CPU_SYS_MODE_UTIL
GBL_CPU_USER_MODE_UTIL
GBL_CPU_SYSCALL_UTIL
GBL_CPU_INTERRUPT_UTIL
GBL_PRI_QUEUE
GBL_CPU_CSWITCH_UTIL
GBL_SWAP_SPACE_UTIL
GBL_DISK_UTIL_PEAK
GBL_DISK_SUBSYSTEM_QUEUE
GBL_MEM_UTIL
GBL_MEM_CACHE_HIT_PCT
GBL_MEM_PAGEOUT_RATE
GBL_MEM_SWAPOUT_RATE
GBL_MEM_QUEUE
GBL_NET_PACKET_RATE
GBL_NET_OUTQUEUE
GBL_NETWORK_SUBSYSTEM_QUEUE


Rgds...Geoff

[Hector Vargas]

Still following up with HP support.