[jira] [Created] (IGNITE-8383) SQL: CREATE INDEX should iterate over data pages rather than PK index

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

[jira] [Created] (IGNITE-8383) SQL: CREATE INDEX should iterate over data pages rather than PK index

Anton Vinogradov (Jira)
Vladimir Ozerov created IGNITE-8383:
---------------------------------------

             Summary: SQL: CREATE INDEX should iterate over data pages rather than PK index
                 Key: IGNITE-8383
                 URL: https://issues.apache.org/jira/browse/IGNITE-8383
             Project: Ignite
          Issue Type: Task
          Components: sql
    Affects Versions: 2.4
            Reporter: Vladimir Ozerov
             Fix For: 2.6


See {{org.apache.ignite.internal.processors.query.schema.SchemaIndexCacheVisitorImpl#processPartition}} method - currently to create an index we iterate over partition tree. This leads to additional page reads because we need to dereference index entry. Instead it would be more efficient to iterate over data pages directly.

Note that there are several bad scenarios when this approach may not perform well:
1) Huge number of overflow pages
2) Shared cache groups



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)