sysbench-developers team mailing list archive

[Learner <question260409@xxxxxxxxxxxxxxxxxxxxx>]

New question #260409 on sysbench:
https://answers.launchpad.net/sysbench/+question/260409

Hello,

I'm running mySQL between on x86 machine and sysbench from another 10G direct connected x86 server.

I captured wireshark trace of transactions at mySQL server side and notice that some "response execute command"  have 4 to 5KByte of response size, where most of the others have 100 to 200 byte response size....Looking at the request sent, both look quite similar - I'm very new to mySQL, so might have missed something critical but comparing the two requests didn't show much different; whereas response was 100B vs 4000B...

Could someone share what could be going on....aggregating multiple responses would help in improving perf, so would like to know how to go about doing that. 

Thanks a lot for your time and help!

Following is the example of command issued on client side:
sysbench --max-time=120 --max-requests=0 --num-threads=100 --test=oltp --mysql-table-engine=innodb --oltp-num-tables=14  --oltp-table-size=XXXX --mysql-db=sysbench --oltp-point-selects=14 --mysql-user=myuser  --mysql-password=XXXX --mysql-host=1.1.1.1 run

-- 
You received this question notification because you are a member of
sysbench-developers, which is an answer contact for sysbench.