5 LOB Storage

Initializing Persistent LOBs to NULL or Empty

You can set a persistent LOB — ­that is, a LOB column in a table, or a LOB attribute in an object type that you defined— to be NULL or empty:

• Setting a Persistent LOB to NULL: A LOB set to NULL has no locator. A NULL value is stored in the row in the table, not a locator. This is the same process as for all other datatypes.

• Setting a Persistent LOB to Empty: By contrast, an empty LOB stored in a table is a LOB of zero length that has a locator. So, if you SELECT from an empty LOB column or attribute, then you get back a locator which you can use to populate the LOB with data using supported programmatic environments, such as OCI or PL/SQL(DBMS_LOB). See Chapter 7, "Overview of Supplied LOB APIs" for more information on supported environments.

Details for these options are given in the following discussions.

SELECT COUNT (*) FROM print_media WHERE ad_graphic IS NOT NULL; 

SELECT COUNT (*) FROM print_media WHERE ad_graphic IS NULL; 

INSERT INTO a_table VALUES (EMPTY_BLOB());

DECLARE
   Lob_loc  BLOB;
BEGIN
   INSERT INTO a_table VALUES (EMPTY_BLOB()) RETURNING blob_col INTO Lob_loc;
   /* Now use the locator Lob_loc to populate the BLOB with data */
END;

Initializing LOBs

You can initialize the LOBs in print_media by using the following INSERT statement:

INSERT INTO print_media VALUES (1001, EMPTY_CLOB(), EMPTY_CLOB(), NULL,
    EMPTY_BLOB(), EMPTY_BLOB(), NULL, NULL, NULL, NULL);

This sets the value of ad_sourcetext, ad_fltextn, ad_composite, and ad_photo to an empty value, and sets ad_graphic to NULL.

Initializing Persistent LOB Columns and Attributes to a Value

You can initialize the LOB column or LOB attributes to a value that contains more than 4G bytes of data, the limit before release 10.2.

Initializing BFILEs to NULL or a File Name

A BFILE can be initialized to NULL or to a filename. To do so, you can use the BFILENAME() function.

Restriction on First Extent of a LOB Segment

The first extent of any segment requires at least 2 blocks (if FREELIST GROUPS was 0). That is, the initial extent size of the segment should be at least 2 blocks. LOBs segments are different because they need at least 3 blocks in the first extent. If you try to create a LOB segment in a permanent dictionary managed tablespace with initial = 2 blocks, then it will still work because it is possible for segments in permanent dictionary managed tablespaces to override the default storage setting of the tablespaces.

But if uniform locally managed tablespaces or dictionary managed tablespaces of the temporary type, or locally managed temporary tablespaces have an extent size of 2 blocks, then LOB segments cannot be created in these tablespaces. This is because in these tablespace types, extent sizes are fixed and the default storage setting of the tablespaces is not ignored.

LOBs Compared to LONG and LONG RAW Types

Table 5-1 lists the similarities and differences between LOBs, LONGs, and LONG RAW types.

Storing Varying-Width Character Data in LOBs

Varying-width character data in CLOB and NCLOB datatypes is stored in an internal format that is compatible with UCS2 Unicode character set format. This ensures that there is no storage loss of character data in a varying-width format. Also note the following if you are using LOBs to store varying-width character data:

• You can create tables containing CLOB and NCLOB columns even if you use a varying-width CHAR or NCHAR database character set.

• You can create a table containing a datatype that has a CLOB attribute regardless of whether you use a varying-width CHAR database character set.

Implicit Character Set Conversions with LOBs

For CLOB and NCLOB instances used in OCI (Oracle Call Interface), or any of the programmatic environments that access OCI functionality, character set conversions are implicitly performed when translating from one character set to another.

The DBMS_LOB.LOADCLOBFROMFILE API, performs an implicit conversion from binary data to character data when loading to a CLOB or NCLOB. With the exception of DBMS_LOB.LOADCLOBFROMFILE, LOB APIs do not perform implicit conversions from binary data to character data.

For example, when you use the DBMS_LOB.LOADFROMFILE API to populate a CLOB or NCLOB, you are populating the LOB with binary data from a BFILE. In this case, you must perform character set conversions on the BFILE data before calling DBMS_LOB.LOADFROMFILE.

Inline and Out-of-Line LOB Storage

LOB columns store locators that reference the location of the actual LOB value. Depending on the column properties you specify when you create the table, and depending the size of the LOB, actual LOB values are stored either in the table row (inline) or outside of the table row (out-of-line).

LOB values are stored out-of-line when any of the following situations apply:

• If you explicitly specify DISABLE STORAGE IN ROW for the LOB storage clause when you create the table.

• If the size of the LOB is greater than approximately 4000 bytes (4000 minus system control information), regardless of the LOB storage properties for the column.

• If you update a LOB that is stored out-of-line and the resulting LOB is less than approximately 4000 bytes, it is still stored out-of-line.

LOB values are stored inline when any of the following conditions apply:

• When the size of the LOB stored in the given row is small, approximately 4000 bytes or less, and you either explicitly specify ENABLE STORAGE IN ROW or the LOB storage clause when you create the table, or when you do not specify this parameter (which is the default).

• When the LOB value is NULL (regardless of the LOB storage properties for the column).

Using the default LOB storage properties (inline storage) can allow for better database performance; it avoids the overhead of creating and managing out-of-line storage for smaller LOB values. If LOB values stored in your database are frequently small in size, then using inline storage is recommended.

Defining Tablespace and Storage Characteristics for Persistent LOBs

When defining LOBs in a table, you can explicitly indicate the tablespace and storage characteristics for each persistent LOB column.

To create a BASICFILE LOB, the BASICFILE keyword is optional but is recommended for clarity, as shown in the following example:

CREATE TABLE ContainsLOB_tab (n NUMBER, c CLOB)  
      lob (c) STORE AS BASICFILE segname (TABLESPACE lobtbs1 CHUNK 4096 
                        PCTVERSION 5 
                        NOCACHE LOGGING 
                        STORAGE (MAXEXTENTS 5) 
                       ); 

For SECUREFILE LOBs, the SECUREFILE keyword is necessary, as shown in the following example (assuming TABLESPACE lobtbs1 is ASSM):

CREATE TABLE ContainsLOB_tab1 (n NUMBER, c CLOB)
      lob (c) STORE AS SECUREFILE sfsegname (TABLESPACE lobtbs1
                       RETENTION AUTO
                       CACHE LOGGING
                       STORAGE (MAXEXTENTS 5)
                     );

If you need to modify the LOB storage parameters on an existing LOB column, then use the MODIFY LOB clause of the ALTER TABLE statement.

LOB Storage Characteristics for LOB Column or Attribute

LOB storage characteristics that can be specified for a LOB column or a LOB attribute include the following:

• TABLESPACE

• PCTVERSION or RETENTION

  Note that you can specify either PCTVERSION or RETENTION for BASICFILE LOBs, but not both. For SECUREFILE LOBs, only the RETENTION parameter can be specified.

• CACHE/NOCACHE/CACHE READS

• LOGGING/NOLOGGING

• CHUNK

• ENABLE/DISABLE STORAGE IN ROW

• STORAGE

For most users, defaults for these storage characteristics will be sufficient. If you want to fine-tune LOB storage, then consider the following guidelines.

TABLESPACE and LOB Index

Best performance for LOBs can be achieved by specifying storage for LOBs in a tablespace different from the one used for the table that contains the LOB. If many different LOBs will be accessed frequently, then it may also be useful to specify a separate tablespace for each LOB column or attribute in order to reduce device contention.

The LOB index is an internal structure that is strongly associated with LOB storage. This implies that a user may not drop the LOB index and rebuild it.

The system determines which tablespace to use for LOB data and LOB index depending on your specification in the LOB storage clause:

• If you do not specify a tablespace for the LOB data, then the tablespace of the table is used for the LOB data and index.

• If you specify a tablespace for the LOB data, then both the LOB data and index use the tablespace that was specified.

PCTVERSION

When a BASICFILE LOB is modified, a new version of the BASICFILE LOB page is produced in order to support consistent read of prior versions of the BASICFILE LOB value.

PCTVERSION is the percentage of all used BASICFILE LOB data space that can be occupied by old versions of BASICFILE LOB data pages. As soon as old versions of BASICFILE LOB data pages start to occupy more than the PCTVERSION amount of used BASICFILE LOB space, Oracle Database tries to reclaim the old versions and reuse them. In other words, PCTVERSION is the percent of used BASICFILE LOB data blocks that is available for versioning old BASICFILE LOB data.

PCTVERSION has a default of 10 (%), a minimum of 0, and a maximum of 100.

To decide what value PCTVERSION should be set to, consider the following:

• How often BASICFILE LOBs are updated?

• How often the updated BASICFILE LOBs are read?

Table 5-2, "Recommended PCTVERSION Settings" provides some guidelines for determining a suitable PCTVERSION value.

If your application requires several BASICFILE LOB updates concurrent with heavy reads of BASICFILE LOB columns, then consider using a higher value for PCTVERSION, such as 20%.

Setting PCTVERSION to twice the default value allows more free pages to be used for old versions of data pages. Because large queries may require consistent reads of BASICFILE LOB columns, it may be useful to retain old versions of BASICFILE LOB pages. In this case, BASICFILE LOB storage may grow because the database will not reuse free pages aggressively.

If persistent BASICFILE LOB instances in your application are created and written just once and are primarily read-only afterward, then updates are infrequent. In this case, consider using a lower value for PCTVERSION, such as 5% or lower.

The more infrequent and smaller the BASICFILE LOB updates are, the less space must be reserved for old copies of BASICFILE LOB data. If existing BASICFILE LOBs are known to be read-only, then you could safely set PCTVERSION to 0% because there would never be any pages needed for old versions of data.

RETENTION Parameter for BASICFILE LOBS

As an alternative to the PCTVERSION parameter, you can specify the RETENTION parameter in the LOB storage clause of the CREATE TABLE or ALTER TABLE statement. Doing so, configures the LOB column to store old versions of LOB data for a period of time, rather than using a percentage of the table space. For example:

CREATE TABLE ContainsLOB_tab (n NUMBER, c CLOB)  
      lob (c) STORE AS BASICFILE segname (TABLESPACE lobtbs1 CHUNK 4096 
                        RETENTION 
                        NOCACHE LOGGING 
                        STORAGE (MAXEXTENTS 5) 
                       ); 

The RETENTION parameter is designed for use with Undo features of the database, such as Flashback Versions Query. When a LOB column has the RETENTION property set, old versions of the LOB data are retained for the amount of time specified by the UNDO_RETENTION parameter.

Note the following with respect to the RETENTION parameter:

• Undo SQL is not enabled for LOB columns as it is with other datatypes. You must set the RETENTION property on a LOB column to use Undo SQL on LOB data.

• You cannot set the value of the RETENTION parameter explicitly. The amount of time for retention of LOB versions in determined by the UNDO_RETENTION parameter.

• Usage of the RETENTION parameter is only supported in Automatic Undo Management mode. You must configure your table for use with Automatic Undo Management before you can set RETENTION on a LOB column.

• The LOB storage clause can specify RETENTION or PCTVERSION, but not both.

  See Also:

  • Oracle Database Advanced Application Developer's Guide for more information on using flashback features of the database.

  • Oracle Database SQL Language Reference for details on LOB storage clause syntax.

RETENTION Parameter for SECUREFILE LOBs

Specifying the RETENTION parameter for SECUREFILE LOBs indicates that the database will manage consistent read data for the SECUREFILE storage dynamically, taking into account factors such as the UNDO mode of the database.

• Specify MAX if the database is in FLASHBACK mode to limit the size of the LOB UNDO retention in bytes. If you specify MAX, then you must also specify the MAXSIZE clause in the storage_clause.

• Specify MIN if the database is in FLASHBACK mode to limit the UNDO retention duration for the specific LOB segment to n seconds.

• Specify AUTO if you want to retain UNDO sufficient for consistent read purposes only. This is the default.

• Specify NONE if no UNDO is required for either consistent read or flashback purposes.

The default RETENTION for SECUREFILE LOBs is AUTO.

CACHE / NOCACHE / CACHE READS

When creating tables that contain LOBs, use the cache options according to the guidelines in Table 5-3, "When to Use CACHE, NOCACHE, and CACHE READS":

LOGGING / NOLOGGING Parameter for BASICFILE LOBs

[NO]LOGGING has a similar application with regard to using LOBs as it does for other table operations. In the usual case, if the [NO]LOGGING clause is omitted, then this means that neither NOLOGGING nor LOGGING is specified and the logging attribute of the table or table partition defaults to the logging attribute of the tablespace in which it resides.

For LOBs, there is a further alternative depending on how CACHE is stipulated.

• CACHE is specified and [NO]LOGGING clause is omitted. LOGGING is automatically implemented (because you cannot have CACHE NOLOGGING).

• CACHE is not specified and [NO]LOGGING clause is omitted. The process defaults in the same way as it does for tables and partitioned tables. That is, the [NO]LOGGING value is obtained from the tablespace in which the LOB segment resides.

The following issues should also be kept in mind.

LOGGING/NOLOGGING for SECUREFILE LOBs

NOLOGGING or LOGGING has a similar application with regard to using LOBs as it does for other table operations. In the usual case, if the [NO]LOGGING clause is omitted, then this means that neither NOLOGGING nor LOGGING is specified and the logging attribute of the table or table partition defaults to the logging attribute of the tablespace in which it resides.

For SECUREFILE LOBs, there is a further alternative depending on how CACHE is specified:

• CACHE is specified and the LOGGING clause is omitted, then LOGGING is used.

• CACHE is not specified and [NO]LOGGING clause is omitted. Then the process defaults in the same way as it does for tables and partitioned tables. That is, the [NO]LOGGING value is obtained from the tablespace in which the LOB value resides.

The following issues should also be kept in mind.

CHUNK

A chunk is one or more Oracle blocks. You can specify the chunk size for the BASICFILE LOB when creating the table that contains the LOB. This corresponds to the data size used by Oracle Database when accessing or modifying the LOB value. Part of the chunk is used to store system-related information and the rest stores the LOB value. The API you are using has a function that returns the amount of space used in the LOB chunk to store the LOB value. In PL/SQL use DBMS_LOB.GETCHUNKSIZE. In OCI, use OCILobGetChunkSize().

ENABLE or DISABLE STORAGE IN ROW Clause

You use the ENABLE | DISABLE STORAGE IN ROW clause to indicate whether the LOB should be stored inline (in the row) or out-of-line.

The default is ENABLE STORAGE IN ROW.

Guidelines for ENABLE or DISABLE STORAGE IN ROW

The maximum amount of LOB data stored in the row is the maximum VARCHAR2 size (4000). This includes the control information as well as the LOB value. If you indicate that the LOB should be stored in the row, once the LOB value and control information is larger than approximately 4000, then the LOB value is automatically moved out of the row.

This suggests the following guidelines:

The default, ENABLE STORAGE IN ROW, is usually the best choice for the following reasons:

• Small LOBs: If the LOB is small (less than approximately 4000 bytes), then the whole LOB can be read while reading the row without extra disk I/O.

• Large LOBs: If the LOB is big (greater than approximately 4000 bytes), then the control information is still stored in the row if ENABLE STORAGE IN ROW is set, even after moving the LOB data out of the row. This control information could enable us to read the out-of-line LOB data faster.

However, in some cases DISABLE STORAGE IN ROW is a better choice. This is because storing the LOB in the row increases the size of the row. This impacts performance if you are doing a lot of base table processing, such as full table scans, multi-row accesses (range scans), or many UPDATE/SELECT to columns other than the LOB columns.

Using Domain Indexing on LOB Columns

You might be able to improve the performance of queries by building indexes specifically attuned to your domain. Extensibility interfaces provided with the database allow for domain indexing, a framework for implementing such domain specific indexes.

Indexing LOB Columns Using a Text Index

Depending on the nature of the contents of the LOB column, one of the Oracle Text options could also be used for building indexes. For example, if a text document is stored in a CLOB column, then you can build a text index to speed up the performance of text-based queries over the CLOB column.

Function-Based Indexes on LOBs

A function-based index is an index built on an expression. It extends your indexing capabilities beyond indexing on a column. A function-based index increases the variety of ways in which you can access data.

Function-based indexes cannot be built on nested tables or LOB columns. However, you can build function-based indexes on VARRAYs.

Like extensible indexes and domain indexes on LOB columns, function-based indexes are also automatically updated when a DML operation is performed on the LOB column. Function-based indexes are also updated when any extensible index is updated.

Extensible Indexing on LOB Columns

The database provides extensible indexing, a feature which enables you to define new index types as required. This is based on the concept of cooperative indexing where a data cartridge and the database build and maintain indexes for data types such as text and spatial for example, for On-line-Analytical Processing (OLAP).

The cartridge is responsible for defining the index structure, maintaining the index content during load and update operations, and searching the index during query processing. The index structure can be stored in Oracle as heap-organized, or an index-organized table, or externally as an operating system file.

To support this structure, the database provides an indextype. The purpose of an indextype is to enable efficient search and retrieval functions for complex domains such as text, spatial, image, and OLAP by means of a data cartridge. An indextype is analogous to the sorted or bit-mapped index types that are built-in within the Oracle Server. The difference is that an indextype is implemented by the data cartridge developer, whereas the Oracle kernel implements built-in indexes. Once a new indextype has been implemented by a data cartridge developer, end users of the data cartridge can use it just as they would built-in indextypes.

When the database system handles the physical storage of domain indexes, data cartridges

• Define the format and content of an index. This enables cartridges to define an index structure that can accommodate a complex data object.

• Build, delete, and update a domain index. The cartridge handles building and maintaining the index structures. Note that this is a significant departure from the medicine indexing features provided for simple SQL data types. Also, because an index is modeled as a collection of tuples, in-place updating is directly supported.

• Access and interpret the content of an index. This capability enables the data cartridge to become an integral component of query processing. That is, the content-related clauses for database queries are handled by the data cartridge.

By supporting extensible indexes, the database significantly reduces the effort needed to develop high-performance solutions that access complex datatypes such as LOBs.

Oracle Text Indexing Support for XML

You can create Oracle Text indexes on CLOB columns and perform queries on XML data.

Partitioning a Table Containing LOB Columns

LOBs are supported in RANGE partitioned, LIST partitioned, and HASH partitioned tables. Composite heap-organized tables can also have LOBs.

You can partition a table containing LOB columns using the following techniques:

• When the table is created using the PARTITION BY ... clause of the CREATE TABLE statement.

• Adding a partition to an existing table using the ALTER TABLE ... ADD PARTITION clause.

• Exchanging partitions with a table that already has partitioned LOB columns using the ALTER TABLE ... EXCHANGE PARTITION clause. Note that EXCHANGE PARTITION can only be used when both tables have the same storage attributes, for example, both tables store LOBs out-of-line.

Creating LOB partitions at the same time you create the table (in the CREATE TABLE statement) is recommended. If you create partitions on a LOB column when the table is created, then the column can hold LOBs stored either inline or out-of-line LOBs.

After a table is created, new LOB partitions can only be created on LOB columns that are stored out-of-line. Also, partition maintenance operations, SPLIT PARTITION and MERGE PARTITIONS, will only work on LOB columns that store LOBs out-of-line.

Note that once a table is created, storage attributes cannot be changed. See "LOB Storage Parameters" for more information about LOB storage attributes.

Creating an Index on a Table Containing Partitioned LOB Columns

To improve the performance of queries, you can create indexes on partitioned LOB columns. For example:

CREATE INDEX index_name 
   ON table_name (LOB_column_1, LOB_column_2, ...) LOCAL;

Note that only domain and function-based indexes are supported on LOB columns. Other types of indexes, such as unique indexes are not supported with LOBs.

Moving Partitions Containing LOBs

You can move a LOB partition into a different tablespace. This is useful if the tablespace is no longer large enough to hold the partition. To do so, use the ALTER TABLE ... MOVE PARTITION clause. For example:

ALTER TABLE current_table MOVE PARTITION partition_name 
   TABLESPACE destination_table_space
   LOB (column_name) STORE AS (TABLESPACE current_tablespace);

Splitting Partitions Containing LOBs

You can split a partition containing LOBs into two equally sized partitions using the ALTER TABLE ... SPLIT PARTITION clause. Doing so permits you to place one or both new partitions in a new tablespace. For example:

ALTER TABLE table_name SPLIT PARTITION partition_name
   AT (partition_range_upper_bound)
   INTO (PARTITION partition_name, 
      PARTITION new_partition_name TABLESPACE new_tablespace_name
         LOB (column_name) STORE AS (TABLESPACE tablespace_name)
         ... ;

Merging Partitions Containing LOBs

You can merge partitions that contain LOB columns using the ALTER TABLE ... MERGE PARTITIONS clause. This technique is useful for reclaiming unused partition space. For example:

ALTER TABLE table_name 
   MERGE PARTITIONS partition_1, partition_2 
   INTO PARTITION new_partition TABLESPACE new_tablespace_name
      LOB (column_name) store as (TABLESPACE tablespace_name)
     ... ;