Final Sprint

[KoalaGeo]

Final tasks

• Fix flyway/gitlab-ci so it’s not updating/running too often
• Document changes to WAL log configuration to reduce log size (and implement)
• Fix error ` ERROR: ST_AsMVT: Compiled without protobuf-c support
• Create juypter notebooks which connect to our data and DASSH/CEFAS,with some data processing/analysis examples, including visualization
• [ ]

Stretch

• Next/Previous response - next and previous response doesn't match OGCAPI specs geopython/pygeoapi#990
  • Fix paging bug in pygeoapi where next/previous links point to HTML and not JSON
• Add Vector Tiles from PostGres in pygeoapi / make possible to use pg_tileserv (Add metadata path property for Tiles provider to config geopython/pygeoapi#820)
  • https://www.crunchydata.com/blog/dynamic-vector-tiles-from-postgis
  • https://github.com/pramsey/minimal-mvt
  • https://github.com/CrunchyData/pg_tileserv
• Try to get one EDR endpoint - pygeoapi with postgres provider? Not possible in budget

Other orgs EDR endpoints:

• https://gitserver.mba.ac.uk/external/data_team/edr/
• https://www.dassh.ac.uk/edr/
• https://data-api.cefas.co.uk/index.html?urls.primaryName=Cefas%20OGC%20Environmental%20Data%20Retrieval%20API#/Collection%20metadata

[volcan01010]

I merged the refresh views task. I was then able to run the scripts in pgAdmin.

I also made a new task to document changes to make the WAL log collection to make it take up less space. We should record this information on the README and discuss with Martin.

During notebook creation, Colin found a bug in pygeoapi where the links in prev/next point to the HTML pages, even if you made the request in JSON. I've added that to the todo.

I'm not sure what to do with the failing build pipelines. Maybe we can look at that on Friday, Edd, if you haven't got it working before then. If you can use your customised GitHub Actions YML to get one build running then you could at least us it to deploy a container into the Dev cluster to build a notebook against.