Search results for query: *

hi everyone Our application have a performance problem the application have about 70 thread ,we find that the application used more than 90% cpu (P5 570 whit 8 CPU), but some days ago the application only use 40%cpu. This is the report whit CURT tools (trace 20 second), with this report we find...

Find out the top 10 process using more monery ps aux | sort -nr +4 | head -10

We have an IBM P550 with AIX 5.2-06. These days I find that ent0's data throughput much more than normal (ent0 si the service ethernet card) I want to know how can I trace or view the content of this throughput . I can do this by some tools under windows flatform , but under windows...

Check out "filemon". It will tell you everything you want to know about what is going on with I/O on your system. --But whit the trace report only I can only find out the busy disk and lv. The report file not include process infomation.

thanks a lot

spamly: Thanks a lot. I had use nmon tool to get more infomation about my processes. Other question, can you tell me about the concept of the "IO Wait" ?

My system is IBM P570 whit oracle 9i RAC I find that the “IO WAIT” always above 40% whit topas and many disks very busy for reading and writing. But I can't find out which process make the "IO WAIT" so high. Who can give me some suggestion?

I suggest you tune your "maxperm" parameter whit this command /usr/samples/kernel/vmtune -P10 -p 5 -h1 -t 10 to reduce memory for filesystem cache.

p5wizard : thanks a lot. And I had do what you said this is the result System Summary -------------- processing percent percent total time total time busy time (msec) (incl. idle) (excl. idle) processing category ===========...

some process running on our server using more cpu. this is topas's report Kernel 26.3 |######## User 12.7 |#### Wait 0.0 | Idle 60.9 |################## the process using so much kernel cpu resource...

check your oracle initial file pay attention to the parameter processes=XXX #the max numbers client can logon the oracle server

I think you root filesystem was full. type df -k to check your "/" filesystem % usage

Thanks for everyone's help. The problem was solved. The reason that /tmp filesystem increase so fast was that we start our application by crontab , and in crontab command line we not specified the standard output, so the application use /tmp for it's standard output. and we modified our crontab...

I tried fuser -c /tmp and get some PID, then I tried ps -ef | grep "PID" and find which process is using /tmp ? but how can I get more infomation? and I want to know now is why the processes are using /tmp but whit du -sk /tmp command I only find very small size? I'm sure my program not write...

We have about 50 processes running on AIX 5.2 this processes complied by visualage C++ 5.0 when this processor runing the /tmp filesystem increase very fast. but when we use du command check it we find very small size in /tmp filesystem ? when killed our processes the /tmp decrease to very...

How can I get the %cpu of a process within a time piece without topas and nmon ?

How can I get the %cpu of a process without topas and nmon ?

In topas tools I find some "aioserver" process. Who can tell me what is aioserver?

PLS post more infomation ,such as vmstat , topas ,svmon -G

Hi ALL : I have IBM RS6000 whit AIX version 4.3.3 a process used more than 1.7G real memory . how can I discover why the process use so much memory ?