APNIC’s core registry services are whois, Registration Data Access Protocol (RDAP), Resource Public Key Infrastructure (RPKI), Internet Routing Registry (IRR), and Reverse Domain Name System (rDNS). The availability for these APNIC core services for Q2 2025 is shown in Table 1. Figures from Q1 are included for easy comparison.
| Service | Q1 Availability | Q2 Availability |
| whois and IRR | 99.991% | 99.996% |
| RDAP | 99.998% | 99.998% |
| RPKI | 99.999% | 99.983% |
| rDNS | 99.998% | 99.999% |
Incorporating user-perspective measurements
To measure availability, APNIC has used a third-party monitoring service that sends queries to these services every minute and records downtime when it detects a problem with the service.
One of our 2025 objectives is to include the perspective of the service users in the availability measurements, which has been done for RDAP and RPKI in Q2. At the end of Q3 we will have a full quarter of user-perspective measurements for RDAP and RPKI, and they will be included in the overall availability measurement. IRR, rDNS and whois will be done later this year.
The way we are measuring availability from the user’s perspective is by using what we know from logs of each user request to define if it was successful or not. In RDAP, for example, if the HTTP response code was not an error and if the response was received in less than five seconds then it was successful. The availability as a percentage then is:
100 × successful requests / total requests Similar user-perspective metrics are being identified for each service.
While we have been monitoring error rates of RDAP and RPKI Repository Delta Protocol (RRDP) requests for over a year, RPKI rsync error rates was a new addition. As we are getting started with measuring from the user’s perspective, we will be using weightings to combine the third-party monitoring service availability and the user-perspective availability measurements into an overall availability, with 25% assigned to the user-perspective availability:
Total Availability = 0.75 × third-party availability + 0.25 × user-perspective availabilityThese weightings will be reflected in the availability values we publish in the Q3 edition of this post.
Improving service levels
An important reason for tracking user-perspective availability and performance metrics is so we can react to problems that users are experiencing that may not be detected by our other monitoring systems. We have target thresholds configured for these added user-perspective availability measurements for RDAP and RPKI and our operations team will be taking action when these thresholds aren’t being met.
Having error rates more visible has already helped highlight problems and led to improvements. For example, by increasing the visibility of error rates in RDAP, the team was able to identify that the amount of memory allocated to the application was insufficient. This led to memory allocation increases as short-term fixes, and two rounds of application memory usage optimization and Java Virtual Machine (JVM) tuning to reduce the memory footprint.
Check the service updates page if you experience downtime or errors when accessing any of APNIC’s core services.
The views expressed by the authors of this blog are their own and do not necessarily reflect the views of APNIC. Please note a Code of Conduct applies to this blog.