This
operation is performed for a single query predicate.
Node name: XISCAN
Represents: Its
evaluation narrows down the qualifying return set of rows IDs and
XML node IDs by range scanning any associated index over XML data
before accessing the base table. The use of an index can improve the
performance of a query because the compiler determines whether and
how to use index information to complete the query. This typically
results in:
- The sorting of nodes by document and the elimination
of any duplicates.
- The fetching of each row of the table that
contains the qualifying
documents.
- An XSCAN operation on the XML document.
Performance
suggestions: - Over time, database updates might cause
an index to become fragmented,
resulting in more index pages than necessary. This can be corrected
by dropping and recreating the index, or reorganizing the index.
- When two or more tables are being accessed, access to the inner
table by means of an index can be made more efficient by providing
an index on the join column of the outer table.
- If statistics
are not current, update them using the RUNSTATS command.