Describe the current state/issue
Part of #666. EZID currently uses Matomo for web analytics, which was found to be not performant with the increased traffic following the N2T resolution functionality cut-over.
Describe the desired state/solution
Investigate the viability and potential benefits of implementing a post-processing approach for EZID usage analytics using Apache log files. The investigation should specifically:
Assess the viability of using Matomo (or a similar tool) to parse Apache log files for analytics instead of real-time tracking.
Compare the performance and availability of post-processed log analysis vs. current real-time tracking, especially under high load from resolution requests or similar traffic.
Identify any gaps in functionality or data granularity compared to the current real-time Matomo implementation.
Evaluate the effort required to implement a log parsing solution and integrate it with existing analytics processes and dashboards.
Assess whether post processing would impede real-time troubleshooting or incident handling.
Develop a proof-of-concept implementation to validate findings from the above points.
Describe the current state/issue Part of #666. EZID currently uses Matomo for web analytics, which was found to be not performant with the increased traffic following the N2T resolution functionality cut-over.
Describe the desired state/solution Investigate the viability and potential benefits of implementing a post-processing approach for EZID usage analytics using Apache log files. The investigation should specifically: