Is there any advantage of one scenario vs other as regards
to latency in the connectivity?
Answer: It doesn’t matter from latency perspective. However
scenario 2 is the preferred way to architect especially if the lower layer
doesn’t need to be exposed to public. E.g. do you really want to expose your DB
layer e.g Oracle RAC to public?
Test
Environment:
Standard
A0s
Enhancement:
Collect stats from all hours of the week and all days of the week.
Test
Scenario 1:
[<user>@<VM1> ~]$ sumresponse=0.0; myAverage=0.0; totalCount=50; for
((c=0; c<totalCount; c++)); do response=`nmap -v -Pn <ILB IP> -p <port>
| grep scanned | awk -F in '{print $2}' | awk -F ' ' '{print $1}'`;echo
$response; sumresponse=`echo "$sumresponse + $response" | bc`; sleep
10; done; echo sum $sumresponse; myAverage=`echo "scale = 3; $sumresponse
/ $totalCount" | bc`; echo myAverage $myAverage
0.14
<omitted…>
myAverage
.147
Test
Scenario 2:
[MppUsWest1admin@centosdevrnd
~]$ sumresponse=0.0; myAverage=0.0; totalCount=50; for ((c=0; c<totalCount;
c++)); do response=`nmap -v -Pn <xyz.cloudapp.net>
-p <CS port> | grep scanned | awk -F in '{print $2}' | awk -F ' ' '{print
$1}'`;echo $response; sumresponse=`echo "$sumresponse + $response" |
bc`; sleep 10; done; echo sum $sumresponse; myAverage=`echo "scale = 3;
$sumresponse / $totalCount" | bc`; echo myAverage $myAverage
0.14
<omitted…>
myAverage
.146
No comments:
Post a Comment