Skip to main content
Last updated: 4 Jun 2026

Fix out-of-date search indices

govuk index

GOV.UK content in Search API is stored in the govuk index in Elasticsearch. This is populated from the Publishing API message queue. This can get out of sync with publishing API, which affects any part of the site using it, including navigation pages and related links. This can happen after restoring a backup or reindexing search. When it does, any publish and unpublish messages that have not been processed need to be resent.

To resend the content to the message queue, use the represent_downstream:published_between rake task. This will recover recently published documents that are missing or not up-to-date, and will remove recently unpublished documents.

Run the represent_downstream:published_between['2018-12-17T01:02:30,%202018-12-18T10:20:30'] rake task in Publishing API, but changing the two timestamps to cover the period of downtime.

Other replay options are available, for example replaying all traffic for a single publishing app or doctype. Be aware that these options will replay the entire Publisher API history for that app or doctype, and may take some time.

metasearch index

The metasearch index contains best bets, which used to be published by Search Admin until that functionality was removed in https://github.com/alphagov/search-admin/pull/1174. There is currently no support for reindexing this index. This is known tech debt.

page-traffic index

The page-traffic index contains traffic data from GA4, which is used to update popularity scores on content in the govuk index. (See docs on updating popularity for more information.) There is a nightly cron job, which updates this index. To update it manually, you can run the associated rake task.

Environment syncs

There are nightly cron jobs that synchronise elasticsearch in integration and staging with the production environment. These jobs work in sequence to take a snapshot of a higher level environment (e.g. production) and then restore the snapshot in the lower level environment (e.g. staging). If documents that are published or unpublished in production are not reflected in integration and staging after 1 day, that could be because of issues with the synchronisation jobs. If any part of the workflow needs to be rerun, this can be done manually in Argo:

  1. Login to Argo for the relevant environment
  2. Navigate to the search-index-env-sync app
  3. Find the relevant cron job:
    • search-index-env-sync-green-es6-snapshot takes a snapshot
    • search-index-env-sync-green-es6-cp-prod restores the production snapshot in staging
    • search-index-env-sync-green-es6-cp-stag restores the staging snapshot in integration
  4. Click on the kebab menu next to the job name and click “Create Job”

Taking a snapshot does not cause any downtime. However, please be aware that restoring a snapshop in staging or integration may wipe out changes that other developers are testing.

For more information on environment syncs across GOV.UK see https://docs.publishing.service.gov.uk/manual/govuk-env-sync.html

Related repositories

More in the Search on GOV.UK section

Learn

How to...