The period of sustained throughput should be relative to the total capacity the RSP applicant expects to be able to meet. Put simply, if the RSP applicant states that total system capacity is 10K domains, then the period of sustained throughput would be quite different from a system capable of 10M or 100M domains. RSP applicants should include all the information related to their test methodology and describe their reasoning.
Generally, for performance measurements ICANN would expect to see information such as:
- Methodologies: A description of the methodology for making the capacity estimates (bandwidth & server) as well as the methodologies for creating the test data.
- Bandwidth Capacity: A description of bandwidth capacities and how that correlates to performance capacity.
- Server Capacity: A description of how server capacity is measured, the locations of potential bottlenecks, and how those could be mitigated.
- Measurements: A description of how the RSP applicant measured the server and bandwidth usage during their test, for example what software and/or equipment was used for measuring, e.g. Prometheus/Grafana, Checkmk, Zabbix, etc.