• Post Reply Bookmark Topic Watch Topic
  • New Topic

Load testing on web service  RSS feed

 
Rem Rad
Greenhorn
Posts: 1
  • Likes 1
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
I am performing a load on a java web service.
I deployed the application on 2 different servers.
1 server having 20 CPU and 16GB memory and another server having 8 CPU and 8 GB memory.
I was able to reach 140 tps with the server having 8 CPU and only 100 tps with the server having 20 CPU. My target is to reach more tps than this result.
I monitored the CPU and memory. Both servers having 60% idle CPU.
I monitored the IO on hard disk and there is no issue.

Can you please help me why the result is varying from a server to another and what should I monitor also .

Thank you very much.
 
Claude Moore
Ranch Hand
Posts: 911
9
IBM DB2 Java Netbeans IDE
  • Likes 1
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
What about performances at network I/O level ?
Some times ago I ran into some troubles with an application installed in two different virtual machines: these VMs very identical in terms of CPU, memory, memory reservetion, but they were provided with two different NIC adapter - of course NICs weren't physical ones.
Since my application runs a lot of SQL statements, differences between NICs were noticeable in the middle term.
 
Vijitha Kumara
Bartender
Posts: 4001
42
Chrome Fedora Hibernate
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Also you might need to see how each server utilizes the cores available to the application with increasing load multiple times to identify any pattern(s). Or if there any improvement for the code to be done to improve this...
 
  • Post Reply Bookmark Topic Watch Topic
  • New Topic
Boost this thread!