Appendix C - Perfomance Samples
The hosts used for testing were of the type(https://docs.microsoft.com/en-us/azure/cloud-services/cloud-services-sizes-specs#dv3-series) (Azure), with the following settings:
|vCPU||Memory: GiB||Temp storage (SSD) GiB||Max data disks||Max cached and temp storage throughput: IOPS / MBps (cache size in GiB)||Max uncached disk throughput: IOPS / MBps||Max NICs / Expected network bandwidth (Mbps)|
|8||32||64||16||16000 / 128 (200)||12800 / 192||4 / 4000|
The same tests were run with two different nevisProxy configurations. One nevisProxy configuration used a local session store, the other one used a remote session store.
Local session store
The first nevisProxy setup contained a LocalSessionStoreServlet to save the sessions. In this configuration, a loadbalancer needs to make sure that session stickiness is enabled. This means that each request for the same session has to be sent to the same nevisProxy instance. ".
Remote session store
The second nevisProxy setup included a MysqlSessionStoreServlet (using replication) to save the sessions. This setup works without the need of session stickiness, but it is slower due to the fact that the data has to be saved into a database. ".
|Config||#Hosts||#Instances per host||Requests/sec||Response time (ms)||Maximum concurrent users|
|Local Session Store||4||1||750||15||10|
|Remote Session Store||4||1||750||400||150|
The system configuration and infrastructure affect the performance. To measure the real performance, running load tests on the target environment is therefore recommended. If you have significantly different results on your system or if you require higher performance, get in contact with the product team for a consult.