Questions about COSBench Final Result (it is different from Performance Graph)

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Questions about COSBench Final Result (it is different from Performance Graph)

wanglixue
Hi,administrator

i have lots of questions about the cosbench Final Result . it has big difference with the Performance Graph.

like below:(sorry , i can not upload my image data)

op1:prepare-write :

op-count  byte-count avg-resTime  avg-ProcTime Throughput Bandwidth

400kops   25.6GB      176.34ms     175.55ms       1363.8 op/s     87.28 MB/S

in COSbenchUserGuide ,
Throughput = op-count  / TotalTime = 400kops / TotalTime
Bandwidth  = byte-count / TotalTime = 25.6GB /TotalTime

according to the result , TotalTime  = 293 S .

BUT, BUT,BUT,BUT

in Performance Graph ,we can see the time is 480 .

why there is so  big difference ??????

can you describe the detail  computational formula ?

thank you very much !!!!


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

RE: Questions about COSBench Final Result (it is different from Performance Graph)

ywang19
Administrator

Can you paste the workload file? for what stage the 480 seconds?

 

 

 

From: wanglixue [via COSBench] [mailto:ml-node+[hidden email]]
Sent: Thursday, July 07, 2016 10:50 AM
To: Wang, Yaguang <[hidden email]>
Subject: Questions about COSBench Final Result (it is different from Performance Graph)

 

Hi,administrator

i have lots of questions about the cosbench Final Result . it has big difference with the Performance Graph.

like below:(sorry , i can not upload my image data)

op1:prepare-write :

op-count  byte-count avg-resTime  avg-ProcTime Throughput Bandwidth

400kops   25.6GB      176.34ms     175.55ms       1363.8 op/s     87.28 MB/S

in COSbenchUserGuide ,
Throughput = op-count  / TotalTime = 400kops / TotalTime
Bandwidth  = byte-count / TotalTime = 25.6GB /TotalTime

according to the result , TotalTime  = 293 S .

BUT, BUT,BUT,BUT

in Performance Graph ,we can see the time is 480 .

why there is so  big difference ??????

can you describe the detail  computational formula ?

thank you very much !!!!




To start a new topic under COSBench, email [hidden email]
To unsubscribe from COSBench, click here.
NAML

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

RE: Questions about COSBench Final Result (it is different from Performance Graph)

wanglixue
can you give me a email address?

i can email to you the detail information.

for my company deny upload image and file .


thank you !!!
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

RE: Questions about COSBench Final Result (it is different from Performance Graph)

wanglixue
In reply to this post by ywang19
Dear ywang19

this below is another example.it has all the detail information.

1、the workload file content:

<?xml version="1.0" encoding="UTF-8"?>
<workload name="test-read4K" description="sample workload configuration">
    <auth type="none" config=""/>
    <storage type="s3" config="accesskey=C9A30CEBC3468BB0F82B;secretkey=6bp9OC2RyE+m+7fnlcdH4p7amIwAAAFVw0aLsGgG;endpoint=http://16.100.54.4:5080/"/>
    <workflow>
        <workstage name="prepare" closuredelay="0">
            <auth type="none" config=""/>
            <storage type="s3" config="accesskey=C9A30CEBC3468BB0F82B;secretkey=6bp9OC2RyE+m+7fnlcdH4p7amIwAAAFVw0aLsGgG;endpoint=http://16.100.54.4:5080/"/>
            <work name="prepare" type="prepare" workers="200"
                interval="10" division="object" runtime="0" rampup="0"
                rampdown="0" afr="200000" totalOps="100" totalBytes="0" config="cprefix=222-64-00-;containers=r(701,800);objects=r(1,8000);sizes=c(64)KB;">
                <auth type="none" config=""/>
                <storage type="s3" config="accesskey=C9A30CEBC3468BB0F82B;secretkey=6bp9OC2RyE+m+7fnlcdH4p7amIwAAAFVw0aLsGgG;endpoint=http://16.100.54.4:5080/"/>
                <operation type="prepare" ratio="100" division="object"
                    config="createContainer=false;cprefix=222-64-00-;containers=r(701,800);objects=r(1,8000);sizes=c(64)KB;" id="op1"/>
            </work>
        </workstage>
    </workflow>
</workload>

2、the Final Result is :
the Throughout is 972.55 ops
the op-count is 800K ops



3、the Performance result is 578.45




why the result of the step 2 and step 3 is different, and have big difference ???
Loading...