Uploaded image for project: 'OpenNMS'
  1. OpenNMS
  2. NMS-13865

Lunr search indexing runs out of memory while building the docs

    XMLWordPrintable

Details

    • Bug
    • Status: Resolved (View Workflow)
    • Major
    • Resolution: Fixed
    • Infrastructure
    • None
    • Documentation
    • Security Level: Default (Default Security Scheme)
    • None
    • DB

    Description

      While building our nightly documentation set we see build failures running out of memory with the JavaScript application. The Lunr indexer exceeds the X2-large memory of 16 GB RAM with our large documentation set mainly for Horizon and Meridian.

      There are several ways to mitigate this issue:

      • Migrating away from a local index file strategy from Lunr to a more scalable external doc search solution, e.g. Algolia
      • Keep Lunr local and reduce the number of documentation sets especially for Horizon

      In short term, we could reduce the documentation set for Horizon if would just publish the latest released version we actually support in our community. EOL doc versions, e.g. Horizon < 29 and Meridian < 2019 would be available in the archived/eol docs on vault.opennms.com.

      For Meridian, we would publish the latest version for each major version which is TOG supported for the last 3 years.

      We have mainly bugfix or do optimization within a major version. The documentation set is still valid and there are no differences in documentation set between Meridian 2021.1.7 and 2021.1.8. We would have also another way of showing people a) what are the latest versions we support either Horizon or Meridian and b) what versions of Meridian are active in support and which ones are EOLed.

      Attachments

        Issue Links

          Activity

            People

              indigo Ronny Trommer
              indigo Ronny Trommer
              Votes:
              0 Vote for this issue
              Watchers:
              1 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: