Forum Discussion

sachin_80710's avatar
sachin_80710
Icon for Nimbostratus rankNimbostratus
Nov 19, 2014

LTM Virtual Server rps Scaling issue

Hi team,

 

One of our customer is having LTM VS rps scaling issue. Client system is a tool that generates request. Few months back we have done similar test on other device and we got good result example 3 client generating request and 3 servers in pools, total rps was 1 million +. But we are doing same test with almost similar configuration on different device we are getting max. 49000 rps. Initially we though could be hardware issue, for check this we forced active device to standby and did similar test on other device in HA. But still we got similar result 49000. Hardware viprion 4400 with two blades. we tired even adding more server behind VS and load generating client but still rps didnt cross 53000

 

Any idea how to troubleshoot this issue, what could be the reason that LTM VS is not able to scale it rps.

 

application delivery
deployment
LTM
performance

4 Replies

Sort By
  • What_Lies_Bene1's avatar
    What_Lies_Bene1
    Icon for Cirrostratus rankCirrostratus
    Nov 19, 2014

    RPS being requests per second? If so, why do you suspect the F5? Is the CPU or memory maxing out? Are errors being reported?

     

  • What_Lies_Bene1's avatar
    What_Lies_Bene1
    Icon for Cirrostratus rankCirrostratus
    Nov 19, 2014

    So, are you using the same test servers and clients? The same software version etc. etc. I'm not sure you are looking in the right place, the F5 can handle millions of connections without breaking a sweat.

     

  • What_Lies_Bene1's avatar
    What_Lies_Bene1
    Icon for Cirrostratus rankCirrostratus
    Nov 19, 2014

    So, are you using the same test servers and clients? The same software version etc. etc. I'm not sure you are looking in the right place, the F5 can handle millions of connections without breaking a sweat.