Skip to content

Latest commit

 

History

History
132 lines (95 loc) · 4.2 KB

admin-tasks.md

File metadata and controls

132 lines (95 loc) · 4.2 KB
Raw

Admin Tasks

This is a place to list all of the admin tasks available in Publishing API.

Discarding drafts

If you need to discard a draft of a document, run the discard_draft rake task:

bundle exec rake discard_draft['some-content-id']

Representing data downstream

Sometimes you need to re-send content to the Content Store to ensure consistency.

The following tasks will allow you to specify which content items/editions to add to the queue that updates the Content Store.

  • Represent all editions downstream
bundle exec rake represent_downstream:all

N.B. This task will take several hours and should be used with caution.

  • Represent downstream for a specific document_type
bundle exec rake represent_downstream:document_type['a-document-type']
  • Represent downstream for a rendering application
bundle exec rake represent_downstream:rendering_app['application-name']
  • Represent downstream for a publishing application
bundle exec rake represent_downstream:publishing_app['application-name']
  • Represent downstream content which has at least one link of type taxon
bundle exec rake represent_downstream:tagged_to_taxon
  • Represent an individual edition downstream
bundle exec rake represent_downstream:content_id['some-content-id']
  • Represent an individual edition downstream via the high priority queue
bundle exec rake represent_downstream:high_priority:content_id['some-content-id']
  • Represent multiple editions downstream
bundle exec rake represent_downstream:content_id['some-content-id some-other-content-id']
  • Represent multiple editions downstream via the high priority queue
bundle exec rake represent_downstream:high_priority:content_id['some-content-id some-other-content-id']

N.B. The content ids are separated by a space.

  • Represent documents by document type(s) downstream via the high priority queue
bundle exec rake represent_downstream:high_priority:document_type['a-document-type a-document-type']

N.B. The document types are separated by a space. This has also been set up as a Jenkins job which will need to be enabled.

Populating expanded links into database

The expanded-links endpoint defaults to accessing denormalised data stored in the database. If there are reasons that this is now out of sync or data is missing you can populate it with a couple of rake commands.

It will also be rebuilt any time a piece of content is represented downstream.

  • To populate every document (this will take a long time - hours)
bundle exec rake expanded_links:populate
  • To populate every document of a document_type
bundle exec rake expanded_links:populate_by_document_type['document-type']
  • To purge the expanded links cache
bundle exec rake expanded_links:truncate

Generating CSV reports of publishings and unpublishing by date range

There are two tasks provided which create CSV reports of publishings and unpublishings within a given time range. Both task require a from timestamp argument and a until timestamp argument.

To see all the editions that were published between two times run:

bundle exec rake "csv_report:publishings_by_date_range['2020-10-01 10:00', '2020-12-31 10:00']"

To see all the editions that were unpublished between two times run:

bundle exec rake "csv_report:unpublishings_by_date_range['2020-10-01 10:00', '2020-12-31 10:00']"

Viewing the Sidekiq UI

We have access to the Sidekiq UI but because Publishing API doesn't have a frontend we have to use port forwarding to see it in our live environments.

You'll need to have access to our EKS clusters before you can follow these instructions. There's documentation here on how to do that. This means that you'll need full production access before you can view the Sidekiq UI.

To view the UI run:

kubectl -n apps port-forward deployment/publishing-api 8080:8080

and then navigate to localhost:8080/sidekiq