ODD PI 25.4 Objective 1: 🎬 TiTiler-CMR is production ready

Motivation

TiTiler-CMR has the potential to be a powerful tool to visualize data catalogued in CMR, enabling visualization as soon as data is delivered to CMR with zero data preprocessing. However, given TiTiler-CMR tiles the raw data, performance could be improved.

Description

The VEDA ODD will do a performance assessment to understand bottlenecks and then come up with a plan improving performance.

Acceptance Criteria

• Performance assessment
• 1 performance improvement implemented
• Production instance of titiler-cmr (dependency on data services team)

Sub-tasks

• Establish mailing list for titiler-cmr users (Worldview, GIBS, VEDA instances) and document / share a process for announcing new deployments + features
• Integrate titiler-cmr timeseries into the VEDA UI (dependency on UI team)
• Complete performance assessment and implement improvements:
  • Edge caching via Cloudfront
  • Share findings with other VEDA teams
• Coordinate a production deployment: Staging and production deployment of titiler-cmr for phased deployment of new features (MCP deployment would require data services)
• Work with data services on a monitoring solution (we should be alerted when something like access breaks)
• Support datasets (HLS in Worldview, MiCASA (GHG), TROPESS (AQ), and GPM IMERG (Earthdata VEDA))
• As needed, document how to integrate a dataset into VEDA UI using titiler-cmr. Include details of how to report issues.
• Stretch - Interactive tool to help troubleshoot datasets