Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

We have a situation as follows:

  1. We have a two different virtual machines (VMs) on some remote server farm. The machines are identical in terms of hardware/software(OS) configurations.
  2. We have a J2EE application running on JBoss on each of those two machines.
  3. These two applications are of different version sav V1 on VM1 and V2 on VM2.
  4. We observed some degraded response time for application V2 when accessed via public URL.
  5. When we accessed the application through a secured VPN, there is hardly any difference.
  6. The bandwidth test (upload/download speed, ping etc) shows that VM1 is responding better when accessed via secured VPN.
  7. We concluded that the application does not seem to have performance issue. Because, it that's the case the performance degradation should also be there when access via VPN. So we concluded its the network problem.
  8. But since those two identical VMs are on same network we are looking for the reasons for different responses.

My question is, given the above situation, what could be reasons for such a behavior ?

share|improve this question
2  
There's likely to be a network performance issue, probably due to an overload of asterisks. –  womble Apr 5 '12 at 8:58
    
If you want a 100% consistent behavior, don't use shared environments. –  3molo Apr 5 '12 at 9:05
    
Points 1 and 3 are at odds with each other. They can't be identical if they're running different versions of the appliocation (which is software). –  joeqwerty Apr 5 '12 at 9:48
    
Regardless of whether the VMs are the same, you state early on that the applications are not. –  adaptr Apr 5 '12 at 9:48

1 Answer 1

You must first check on the network level from the OS. Use iperf on to test network latency and bandwidth. It uses only bit in memory to send random traffic. This is used to stress and compare network purely. If iperf to and from both VM are identical, then it comes from the application level (or other things related, like SGBD). iperf work with one acting like server and the other acting as client. Test can be UDP & TCP. Use TCP to be close to your application (web server).

Maybe the hypervisor hosting the bad VM have problem with it's network card, like speed & duplex, or interfaces errors.

VPN make the comparing false, because as VPN add latency and slow down speed, the difference looks slighter (while it's not, as you notice it through normal lan).

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.