[DBWG] Failover Test Maintenance for RDAP Service - Monday 8th

[Manhal]

Dear Colleagues,

We are pleased to announce that the recent failover maintenance test was a success! Access to  our RDAP systems seamlessly transitioned between different sites without any service interruption, even while running at half capacity with only two out of four nodes active during the test.

This comprehensive test focused on evaluating the availability and reachability of our RDAP service and we are now confident that a major outage on any of our primary sites will not affect the availability of the RDAP for our members, and the community.

Thanks for your patience and support.

Regards,
The Infrastructure & Security department.

PS: for techies, a bit more details: the aim of this exercise was to test the transitioning between various sites, including switching from DNS round-robin routing to the JNB Isando site (ISO) load balancer, then from ISO to the JNB Parklands site (PKL), PKL to the Mauritius Telecom site (MRU), and finally returning to our primary site, ISO. This was successful, and you will find below the average response times from different places worldwide, as monitored by Site24x7 for services accessed through each site:

MRU site: 1303.888889 milliseconds
PKL site: 1205.294118 milliseconds
ISO site: 1238.722222 milliseconds



> On 5 Apr 2024, at 10:38, Manhal <manhal.mohammed at afrinic.net> wrote:
> 
> Date: Monday 8th April 2024
> Start Time: 10:00 (GMT+4)
> End Time:   13:00 (GMT+4)
> Duration:   3 hours
> Affected Service: RDAP
> Details:
> Dear Colleagues,
> Please note that on the Monday 8th , 10:00 AM  we plan to test the disaster recovery ability of our RDAP service. The test will consist in switching some servers off on a primary site and check that the traffic is temporarily redirected to an alternate location before being switched back to the primary site.
> The goal of this test is to validate our load-balancing systems, in order to mitigate disruption of critical services delivery to the greatest degree possible if a disruption due to disaster occurs at all in the future.
> Note that during and after the test, we anticipate no disruptions in service accessibility, with DNS changes propagating within 60 seconds. Should you experience any issues during the test period, please let us know immediately so that we can take all the necessary remedial actions.
> 
> 
> Regards,
> Infrastructure & Security department

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.afrinic.net/pipermail/dbwg/attachments/20240408/8df64c49/attachment.html>