GPL-988 investigate queries required for pipeline kanban

[KatyTaylor]

Description Look into what queries (API calls) would be needed for a kanban style display of a pipeline within Limber.

Who the primary contacts are for this work Katy, James G, Tom W

Additional context or information This is part of the wider requirements for having an overview of what plates are where / reporting / replacing the Heron Trello board. (see https://github.com/sanger/General-Backlog-Items/issues/139)

[TWJW-SANGER]

Link to UKB - Heron WIP meeting notes and agreements

[KatyTaylor]

Used these options:

• Sequencescape V2 API (JSON API)
• join to state_changes (to avoid using transfer requests to determine the state of the plate / tube)
• use 'sparse fieldsets' to avoid unnecessary queries retrieving fields we don't want
• use maximum page size Sequencescape allows (500) as it reduces the number of API requests we have to make, and massively reduces loading time

This gives a loading time of 30 seconds for a proof of concept page, which lists out all plates and tubes of the LTHR-384 Heron pipeline purposes, grouped by purpose, displaying their barcode and state (542 records at time of writing).

This required modifications as seen in these commits:

• Sequencescape:
  • https://github.com/sanger/sequencescape/commit/ffd4a42e7cd223b235384474f4d8428624e7b294
• Limber:
  • https://github.com/sanger/limber/commit/258034dcebc0d10440382c716d0b97da607f9494
  • https://github.com/sanger/limber/commit/dd656c0235d91187bc46632e56dd6bba0bda41bd

[KatyTaylor]

Putting it in code review in case anyone wants to have an early look at the code linked to in previous message. It will be incorporated into the code base as part of the actual kanban story - https://github.com/sanger/General-Backlog-Items/issues/139

[KatyTaylor]

Closing as is in 'Done' column