With the increase in utilization of the CSN, it was decided to establish additional mechanisms that would allow a more resilient day-to-day usage of the CSN. The system was performing well but with no guarantees in respect of information loss. Since event propagation is considered a highly important concept in the normal operation of the CSN, message loss in that area was established as a major risk. For that reason, the team decided to apply additional mechanisms for minimizing that risk. In particular, the innovations that were introduced were:
Support for automatic recovery in the CSN web and worker nodes and the browser plugin.
Support for publisher confirms when sending messages from the web node.
Additional remote RabbitMQ instance for the purpose of disaster recovery. The CSN web and worker nodes and the browser plugin were enhanced to take the remote instance into consideration upon automatic recovery (by extending those nodes with the...