What is a cluster table?

Provided By: John Kazerooni

 

A cluster is a schema object that contains one or more tables that all have one or more columns in common. Rows of one or more tables that share the same value in these common columns are physically stored together within the database.

 

Generally, you should only cluster tables that are frequently joined on the cluster key columns in SQL statements. Clustering multiple tables improves the performance of joins, but it is likely to reduce the performance of full table scans, INSERT statements, and UPDATE statements that modify cluster key values. Before clustering, consider its

 

Cluster Keys

The columns defined by the CREATE CLUSTER command make up the cluster key. These cluster columns must correspond in both datatype and size to columns in each of the clustered tables, although they need not correspond in name.

 

Types of Clusters

A cluster can be either an indexed cluster or a hash cluster.

 

Indexed Clusters

In an indexed cluster, Oracle stores together rows having the same cluster key value. Each distinct cluster key value is stored only once in each data block, regardless of the number of tables and rows in which it occurs. This saves disk space and improves performance for many operations.

You may want to use indexed clusters in the following cases:

·        Your queries retrieve rows over a range of cluster key values.

·        Your clustered tables may grow unpredictably.

After you create an indexed cluster, you must create an index on the cluster key before you can issue any data manipulation language (DML) statements against a table in the cluster. This index is called the cluster index.

A cluster index provides quick access to rows within a cluster based on the cluster key. If you issue a SQL statement that searches for a row in the cluster based on its cluster key value, Oracle searches the cluster index for the cluster key value and then locates the row in the cluster based on its ROWID.

 

Hash Clusters

In a hash cluster, Oracle stores together rows that have the same hash key value. The hash value for a row is the value returned by the cluster's hash function. When you create a hash cluster, you can either specify a hash function or use the Oracle internal hash function. Hash values are not actually stored in the cluster, although cluster key values are stored for every row in the cluster.

You may want to use hash clusters in the following cases:

·        Your queries retrieve rows based on equality conditions involving all cluster key columns.

·        Your clustered tables are static or you can determine the maximum number of rows and the maximum amount of space required by the cluster when you create the cluster.

The hash function provides access to rows in the table based on the cluster key value. If you issue a SQL statement that locates a row in the cluster based on its cluster key value, Oracle applies the hash function to the given cluster key value and uses the resulting hash value to locate the matching rows. Because multiple cluster key values can map to the same hash value, Oracle must also check the row's cluster key value. This process often results in less I/O than the process for the indexed cluster, because the index search is not required.

Oracle's internal hash function returns values ranging from 0 to the value of HASHKEYS - 1. If you specify a column with the HASH IS clause, the column values need not fall into this range. Oracle divides the column value by the HASHKEYS value and uses the remainder as the hash value. The hash value for null is HASHKEYS - 1. Oracle also rounds the HASHKEYS value up to the nearest prime number to obtain the actual number of hash values. This rounding reduces the likelihood of hash collisions, or multiple cluster key values having the same hash value.

You cannot create a cluster index for a hash cluster, and you need not create an index on a hash cluster key.

If you cannot fit all rows for one hash value into a data block, do not use hash clusters. Performance is very poor in this circumstance because an insert or update of a row in a hash cluster with a size exceeding the data block size fills the block and performs row chaining to contain the rest of the row.

 

Cluster Size

Oracle uses the value of the SIZE parameter to determine the space reserved for rows corresponding to one cluster key value or one hash value. This space then determines the maximum number of cluster or hash values stored in a data block. If the SIZE value is not a divisor of the data block size, Oracle uses the next largest divisor. If the SIZE value is larger than the data block size, Oracle uses the operating system block size, reserving at least one data block per cluster or hash value.

Oracle also considers the length of the cluster key when determining how much space to reserve for the rows having a cluster key value. Larger cluster keys require larger sizes. To see the actual size, query the KEY_SIZE column of the USER_CLUSTERS data dictionary view. This does not apply to hash clusters because hash values are not actually stored in the cluster.

Although the maximum number of cluster and hash key values per data block is fixed on a per `-cluster basis, Oracle does not reserve an equal amount of space for each cluster or hash key value. Varying this space stores data more efficiently, because the data stored per cluster or hash key value is rarely fixed.

A SIZE value smaller than the space needed by the average cluster or hash key value may require the data for one cluster key or hash key value to occupy multiple data blocks. A SIZE value much larger results in wasted space.

When you create a hash cluster, Oracle immediately allocates space for the cluster based on the values of the SIZE and HASHKEYS parameters.

 

Adding Tables to a Cluster

You can add tables to an existing cluster by issuing a CREATE TABLE statement with the CLUSTER clause. A cluster can contain as many as 32 tables, although the performance gains of clustering are often lost in clusters of more than four or five tables.

 

All tables in the cluster have the cluster's storage characteristics as specified by the PCTUSED, PCTFREE, INITRANS, MAXTRANS, TABLESPACE, and STORAGE parameters.

 

Example I

The following statement creates an indexed cluster named PERSONNEL with the cluster key column DEPARTMENT_NUMBER, a cluster size of 512 bytes, and storage parameter values:

CREATE CLUSTER personnel

    ( department_number  NUMBER(2) )

    SIZE 512

    STORAGE (INITIAL 100K NEXT 50K PCTINCREASE 10);

 

The following statements add the EMP and DEPT tables to the cluster:

CREATE TABLE emp

    (empno     NUMBER        PRIMARY KEY,

     ename     VARCHAR2(10)  NOT NULL

                             CHECK (ename = UPPER(ename)),

     job       VARCHAR2(9),

     mgr       NUMBER        REFERENCES scott.emp(empno),

     hiredate  DATE          CHECK (hiredate >= SYSDATE),

     sal       NUMBER(10,2)  CHECK (sal > 500),

     comm      NUMBER(9,0)   DEFAULT NULL,

     deptno   NUMBER(2)      NOT NULL )

     CLUSTER personnel (deptno);

 

CREATE TABLE dept

    (deptno  NUMBER(2),

     dname   VARCHAR2(9),

     loc     VARCHAR2(9))

     CLUSTER personnel (deptno);

 

The following statement creates the cluster index on the cluster key of PERSONNEL:

CREATE INDEX idx_personnel ON CLUSTER personnel;

 

After creating the cluster index, you can insert rows into either the EMP or DEPT tables.

 

Example II

The following statement creates a hash cluster named PERSONNEL with the cluster key column DEPARTMENT_NUMBER, a maximum of 503 hash key values, each of size 512 bytes, and storage parameter values:

CREATE CLUSTER personnel

( department_number  NUMBER )

SIZE 512  HASHKEYS 500

    STORAGE (INITIAL 100K  NEXT 50K  PCTINCREASE 10);

 

Because the above statement omits the HASH IS clause, Oracle uses the internal hash function for the cluster.

 

Example III

The following statement creates a hash cluster named PERSONNEL with the cluster key made up of the columns HOME_AREA_CODE and HOME_PREFIX, and uses a SQL expression containing these columns for the hash function:

CREATE CLUSTER personnel

    ( home_area_code  NUMBER,

    home_prefix     NUMBER )

    HASHKEYS 20

    HASH IS MOD(home_area_code + home_prefix, 101);