Uploaded image for project: 'Couchbase Server'
  1. Couchbase Server
  2. MB-23380

Decrease storage footprint for Array Indexing

    XMLWordPrintable

Details

    • Improvement
    • Resolution: Unresolved
    • Major
    • Neo.next
    • 5.0.0
    • secondary-index
    • None

    Description

      Customers want to store billions of documents with tens or hundreds of items in an array index.

      Our current storage scheme results in a very large index which directly impacts TCO / Hardware Requirements (Memory/Nodes etc.)

      Creating this ticket as a placeholder for tracking the improvement in Spock.Next

       

      As an example for 100 milliion documents, with 100 elements in an array each object of 10 bytes and a DocId size of 20 bytes. If you are using 128 GB of RAM and plan for 20% headroom for growth you are going to need 51 nodes for this single index without HA. This is just non tenable from a TCO standpoint - it excludes us from a lot of usecases

       

      We need to fix this for both MOI and Plasma

      Attachments

        No reviews matched the request. Check your Options in the drop-down menu of this sections header.

        Activity

          People

            prathibha Prathibha Bisarahalli (Inactive)
            asif.kazi Asif Kazi (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            11 Start watching this issue

            Dates

              Created:
              Updated:

              Gerrit Reviews

                There are no open Gerrit changes

                PagerDuty