Skip to content

Latest commit

 

History

History
69 lines (48 loc) · 2.63 KB

README-quicktestcase-24,32,48CPUs.md

File metadata and controls

69 lines (48 loc) · 2.63 KB

Do the following:

  1. Untar the file as oracle user at /home/oracle/dba/cputoolkit

cd /home/oracle/dba/cputoolkit (if the directory does not exist, mkdir -p /home/oracle/dba/cputoolkit) tar -xjvpf cputoolkit.tar.bz2 chmod 755 *

  1. On another terminal window login as oracle then go to /home/oracle/dba/cputoolkit

sh cpu_topology <-- gets the CPU config (outputs cpu_topology.log) sh cage_check.sh <-- gets the Resource Management (RM) & Scheduler config (outputs cage_check.log) sqlplus "/ as sysdba" @disable_rm.sql <-- disables the scheduler windows, RM, & sets CPU_COUNT

  1. Run the benchmark a. If it's a RAC environment and if "dw" is the database name, that's going to be dw1 on the script parameter if you are running it on instance #1 b. If the server has 48 max CPUs you can do the benchmark by 4 CPUs increment, and per increment it will run for 10 minutes c. While the benchmark is running it generates a lot of instrumentation logs (ash, load profile, etc.)

./runcputoolkit-auto ./runcputoolkit-auto 1 4 49 dw1 run for 1 CPU/s run for 5 CPU/s run for 9 CPU/s run for 13 CPU/s run for 17 CPU/s run for 21 CPU/s run for 25 CPU/s run for 29 CPU/s run for 33 CPU/s run for 37 CPU/s run for 41 CPU/s run for 45 CPU/s run for 49 CPU/s

  1. After the run, do this.. basically putting all txt and log files to a folder then just tar it then send it to me

mkdir <testcase_folder_name> mv *txt <testcase_folder_name> mv *log <testcase_folder_name>

Also put the the RM config back by looking at the following files and re-enable everything that has been disabled

disable_rm.log cage_check.log

--re-enable execute dbms_scheduler.set_attribute('MONDAY_WINDOW','RESOURCE_PLAN',''); execute dbms_scheduler.set_attribute('TUESDAY_WINDOW','RESOURCE_PLAN',''); execute dbms_scheduler.set_attribute('WEDNESDAY_WINDOW','RESOURCE_PLAN',''); execute dbms_scheduler.set_attribute('THURSDAY_WINDOW','RESOURCE_PLAN',''); execute dbms_scheduler.set_attribute('FRIDAY_WINDOW','RESOURCE_PLAN',''); execute dbms_scheduler.set_attribute('SATURDAY_WINDOW','RESOURCE_PLAN',''); execute dbms_scheduler.set_attribute('SUNDAY_WINDOW','RESOURCE_PLAN',''); execute DBMS_AUTO_TASK_ADMIN.ENABLE; alter system set resource_manager_plan= scope=both sid='&INST';

  1. Whenever you want to kill the run just execute the die.sh script

sh die.sh

-Karl