Uploaded image for project: 'Nuxeo Platform'
  1. Nuxeo Platform
  2. NXP-17422

Elasticsearch UIDSequencer doesn't use the sequence index contribution

    XMLWordPrintable

    Details

    • Impact type:
      Configuration Change
    • Upgrade notes:
      Hide

      Changed Elasticsearch index name used by the uidgen sequencer from "sequence" to "uidgen"

      Show
      Changed Elasticsearch index name used by the uidgen sequencer from "sequence" to "uidgen"
    • Sprint:
      Sprint Drive 7.2-7
    • Story Points:
      1

      Description

      Indeed, ESUIDSequencer#IDX_NAME = "uidgen" which is different from the "sequence" index name defined in elasticsearch-sequence-index-contrib.xml:

          <elasticSearchIndex name="sequence" type="seqId">
      
            <settings>
              {
              "number_of_shards" : 1,
              "auto_expand_replicas" : "0-all"
              }
            </settings>
      
            <mapping>
              {
              "_source" : {"enabled": false},
              "_all" : {"enabled": false},
              "_type" : {"index": "no"},
              "enabled" : false
              }
            </mapping>
      
          </elasticSearchIndex>
      
        </extension>
      

      In fact, the sequence number generation works fine as ESUIDSequencer#getNext generates the "uidgen" index on the fly if it doesn't exist, but:

      • This is not optimal as we store the source => bad performances.
      • We create an empty "sequence" index.
      • This is simply inconsistent!

        Attachments

          Issue Links

            Activity

              People

              • Votes:
                0 Vote for this issue
                Watchers:
                2 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: