18 Very Large Databases (VLDB)

Partition Key

Each row in a partitioned table is unambiguously assigned to a single partition. The partition key is a set of one or more columns that determines the partition for each row. Oracle Database automatically directs insert, update, and delete operations to the appropriate partition through the use of the partition key. A partition key:

• Consists of an ordered list of 1 to 16 columns

• Cannot contain a LEVEL, ROWID, or MLSLABEL pseudocolumn or a column of type ROWID

• Can contain columns that are NULLable

Partitioned Tables

Tables can be partitioned into up to 1024K-1 separate partitions. Any table can be partitioned except those tables containing columns with LONG or LONG RAW datatypes. You can, however, use tables containing columns with CLOB or BLOB datatypes.

Partitioned Index-Organized Tables

You can partition index-organized tables by range, list, or hash. Partitioned index-organized tables are very useful for providing improved manageability, availability, and performance for index-organized tables. In addition, data cartridges that use index-organized tables can take advantage of the ability to partition their stored data.

For partitioning an index-organized table:

• Partition columns must be a subset of primary key columns

• Secondary indexes can be partitioned—locally and globally

• OVERFLOW data segments are always equipartitioned with the table partitions

Partitioning Methods

Oracle provides the following partitioning methods:

• Range partitioning maps data to partitions based on ranges of partition key values that you establish for each partition.

• Interval partitioning is an extension of range partitioning which instructs the database to automatically create partitions of a specified interval when data inserted into the table exceeds all of the range partitions.

• Hash partitioning maps data to partitions based on a hashing algorithm that evenly distributes rows among partitions, giving partitions approximately the same size.

• List partitioning enables you to explicitly control how rows map to partitions by specifying a list of discrete values in the description for each partition.

• Reference partitioning enables you to partition a table based on the partitioning scheme of the table referenced in its referential constraint.

• Composite partitioning is a combination of two partitioning methods to further divide the data into subpartitions:

  • Composite range-range partitioning partitions data using the range method, and within each partition, subpartitions it using the range method.

  • Composite range-hash partitioning partitions data using the range method, and within each partition, subpartitions it using the hash method.

  • Composite range-list partitioning partitions data using the range method, and within each partition, subpartitions it using the list method.

  • Composite list-range partitioning partitions data using the list method, and within each partition, subpartitions it using the range method.

  • Composite list-hash partitioning partitions data using the list method, and within each partition, subpartitions it using the hash method.

  • Composite list-list partitioning partitions data using the list method, and within each partition, subpartitions it using the list method.

• System partitioning enables application-controlled partitioning for arbitrary tables.

Local Partitioned Indexes

Local partitioned indexes are easier to manage than other types of partitioned indexes. They also offer greater availability and are common in DSS environments. The reason for this is equipartitioning: each partition of a local index is associated with exactly one partition of the table. This enables Oracle Database to automatically keep the index partitions in sync with the table partitions, and makes each table-index pair independent. Any actions that make one partition's data invalid or unavailable only affect a single partition.

Local partitioned indexes support more availability when there are partition or subpartition maintenance operations on the table. A type of index called a local nonprefixed index is very useful for historical databases. In this type of index, the partitioning is not on the left prefix of the index columns.

You cannot explicitly add a partition to a local index. Instead, new partitions are added to local indexes only when you add a partition to the underlying table. Likewise, you cannot explicitly drop a partition from a local index. Instead, local index partitions are dropped only when you drop a partition from the underlying table.

A local index can be unique. However, in order for a local index to be unique, the partitioning key of the table must be part of the index's key columns. Unique local indexes are useful for OLTP environments.

Global Partitioned Indexes

Oracle Database offers two types of global partitioned index: range partitioned and hash partitioned.

This section includes the following topics:

• Global Range Partitioned Indexes

• Global Hash Partitioned Indexes

• Maintenance of Global Partitioned Indexes

ADD (HASH) 
COALESCE (HASH) 
DROP 
EXCHANGE 
MERGE 
MOVE 
SPLIT 
TRUNCATE 

ALTER TABLE DROP PARTITION P1 UPDATE INDEXES;

Global Nonpartitioned Indexes

Global nonpartitioned indexes behave just like a nonpartitioned index. They are commonly used in OLTP environments and offer efficient access to any individual record.

Miscellaneous Information about Creating Indexes on Partitioned Tables

You can create bitmap indexes on partitioned tables, with the restriction that the bitmap indexes must be local to the partitioned table. They cannot be global indexes.

Global indexes can be unique. Local indexes can only be unique if the partitioning key is a part of the index key.

Using Partitioned Indexes in OLTP Applications

Here are a few guidelines for OLTP applications:

• Global indexes and unique, local indexes provide better performance than nonunique local indexes because they minimize the number of index partition probes.

• Local indexes offer better availability when there are partition or subpartition maintenance operations on the table.

• Hash-partitioned global indexes offer better performance by spreading out contention when the index is monotonically growing. In other words, most of the index insertions occur only on the right edge of an index.

Using Partitioned Indexes in Data Warehousing and DSS Applications

Here are a few guidelines for data warehousing and DSS applications:

• Local indexes are preferable because they are easier to manage during data loads and during partition-maintenance operations.

• Local indexes can improve performance because many index partitions can be scanned in parallel by range queries on the index key.

Partitioned Indexes on Composite Partitions

Here are a few points to remember when using partitioned indexes on composite partitions:

• Subpartitioned indexes are always local and stored with the table subpartition by default.

• Tablespaces can be specified at either index or index subpartition levels.

Partition Pruning

Oracle Database explicitly recognizes partitions and subpartitions. It then optimizes SQL statements to mark the partitions or subpartitions that need to be accessed and eliminates (prunes) unnecessary partitions or subpartitions from access by those SQL statements. In other words, partition pruning is the skipping of unnecessary index and data partitions or subpartitions in a query.

For each SQL statement, depending on the selection criteria specified, unneeded partitions or subpartitions can be eliminated. For example, if a query only involves March sales data, then there is no need to retrieve data for the remaining eleven months. Such intelligent pruning can dramatically reduce the data volume, resulting in substantial improvements in query performance.

The SQL Access Advisor can automatically analyze the schema design for a given workload and recommend indexes, function-based indexes, partitions, and materialized views to create, retain, or drop as appropriate for the workload.

If the optimizer determines that the selection criteria used for pruning are satisfied by all the rows in the accessed partition or subpartition, it removes those criteria from the predicate list (WHERE clause) during evaluation in order to improve performance. However, the optimizer cannot prune partitions if the SQL statement applies a function to the partitioning column (with the exception of the TO_DATE function). Similarly, the optimizer cannot use an index if the SQL statement applies a function to the indexed column, unless it is a function-based index.

Pruning can eliminate index partitions even when the underlying table's partitions cannot be eliminated, but only when the index and table are partitioned on different columns. You can often improve the performance of operations on large tables by creating partitioned indexes that reduce the amount of data that your SQL statements need to access or modify.

Equality, range, LIKE, and IN-list predicates are considered for partition pruning with range or list partitioning, and equality and IN-list predicates are considered for partition pruning with hash partitioning.

SELECT SUM(value)
FROM cust_orders
WHERE order_date BETWEEN '28-MAR-98' AND '23-APR-98';

Partition-wise Joins

A partition-wise join is a join optimization for joining two tables that are both partitioned along the join column(s). With partition-wise joins, the join operation is broken into smaller joins that are performed sequentially or in parallel. Another way of looking at partition-wise joins is that they minimize the amount of data exchanged among parallel slaves during the execution of parallel joins by taking into account data distribution.