We had a 3 node cluster running on HDDs. After migrating to SSDs, we ran a load test but haven't seen any latency difference.

So we are thinking about running a stress test. Is there any threshold or parameters we should set or any general recommendation to find the utmost difference?

  • define latency, do you mean if write or in read path ? What is the cluster configuration (with replicas ?) What tests do you conduct and what are the results you observer before and after. – Baptiste Mille-Mathias Aug 7 '16 at 8:25

Maybe you should run a diagnostics test on the cluster itself instead. There might be some unknown entity(ies) that would be preventing the latency difference. Even on a load test, you should see a significant difference.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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