Appendix F. Limits and Restrictions

Table of Contents

F.1. Restrictions on Stored Routines and Triggers
F.2. Restrictions on Server-Side Cursors
F.3. Restrictions on Subqueries
F.4. Restrictions on Views
F.5. Restrictions on XA Transactions
F.6. Restrictions on Character Sets
F.7. Limits in MySQL
F.7.1. Limits of Joins
F.7.2. The Maximum Number of Columns Per Table
F.7.3. Windows Platform Limitations

The discussion here describes restrictions that apply to the use of MySQL features such as subqueries or views.

F.1. Restrictions on Stored Routines and Triggers

Some of the restrictions noted here apply to all stored routines; that is, both to stored procedures and stored functions. Some of restrictions apply only to stored functions, and not to stored procedures.

All of the restrictions for stored functions also apply to triggers.

Stored routines cannot contain arbitrary SQL statements. The following statements are disallowed:

  • The table-maintenance statements CHECK TABLES and OPTIMIZE TABLES. This restriction is lifted beginning with MySQL 5.0.17.

  • The locking statements LOCK TABLES, UNLOCK TABLES.

  • LOAD DATA and LOAD TABLE.

  • SQL prepared statements (PREPARE, EXECUTE, DEALLOCATE PREPARE). Implication: You cannot use dynamic SQL within stored routines (where you construct dynamically statements as strings and then execute them). This restriction is lifted as of MySQL 5.0.13 for stored procedures; it still applies to stored functions and triggers.

    In addition, SQL statements that are not permitted within prepared statements are also not permitted in stored routines. See Section 12.7, “SQL Syntax for Prepared Statements”, for a list of statements supported in prepared statements. Statements not listed there are not supported for SQL prepared statements and thus are also not supported for stored routines unless noted otherwise in Chapter 18, Stored Procedures and Functions.

For stored functions (but not stored procedures), the following additional statements or operations are disallowed:

  • Statements that do explicit or implicit commit or rollback.

  • Statements that return a result set. This includes SELECT statements that do not have an INTO var_list clause and SHOW statements. A function can process a result set either with SELECT ... INTO var_list or by using a cursor and FETCH statements. See Section 18.2.7.3, “SELECT ... INTO Statement”.

  • FLUSH statements.

  • Before MySQL 5.0.10, stored functions created with CREATE FUNCTION must not contain references to tables, with limited exceptions. They may include some SET statements that contain table references, for example SET a:= (SELECT MAX(id) FROM t), and SELECT statements that fetch values directly into variables, for example SELECT i INTO var1 FROM t.

  • Recursive statements. That is, stored functions cannot be used recursively.

  • Within a stored function or trigger, it is not permitted to modify a table that is already being used (for reading or writing) by the statement that invoked the function or trigger.

  • ALTER VIEW.

Note that although some restrictions normally apply to stored functions and triggers but not to stored procedures, those restrictions do apply to stored procedures if they are invoked from within a stored function or trigger. For example, although you can use FLUSH in a stored procedure, such a stored procedure cannot be called from a stored function or trigger.

It is possible for the same identifier to be used for a routine parameter, a local variable, and a table column. Also, the same local variable name can be used in nested blocks. For example:

CREATE PROCEDURE p (i INT)
BEGIN
  DECLARE i INT DEFAULT 0;
  SELECT i FROM t;
  BEGIN
    DECLARE i INT DEFAULT 1;
    SELECT i FROM t;
  END;
END;

In such cases the identifier is ambiguous and the following precedence rules apply:

  • A local variable takes precedence over a routine parameter or table column

  • A routine parameter takes precedence over a table column

  • A local variable in an inner block takes precedence over a local variable in an outer block

The behavior that variables take precedence over table columns is non-standard.

Use of stored routines can cause replication problems. This issue is discussed further in Section 18.5, “Binary Logging of Stored Routines and Triggers”.

INFORMATION_SCHEMA does not have a PARAMETERS table until MySQL 6.0, so applications that need to acquire routine parameter information at runtime must use workarounds such as parsing the output of SHOW CREATE statements or the param_list column of the mysql.proc table. param_list contents can be processed from within a stored routine, unlike the output from SHOW.

There are no stored routine debugging facilities.

Before MySQL 5.0.17, CALL statements cannot be prepared. This true both for server-side prepared statements and for SQL prepared statements.

UNDO handlers are not supported.

FOR loops are not supported.

To prevent problems of interaction between server threads, when a client issues a statement, the server uses a snapshot of routines and triggers available for execution of the statement. That is, the server calculates a list of procedures, functions, and triggers that may be used during execution of the statement, loads them, and then proceeds to execute the statement. This means that while the statement executes, it will not see changes to routines performed by other threads.

For triggers, the following additional statements or operations are disallowed:

  • Triggers currently are not activated by foreign key actions.

  • The RETURN statement is disallowed in triggers, which cannot return a value. To exit a trigger immediately, use the LEAVE statement.

  • Triggers are not allowed on tables in the mysql database.

F.2. Restrictions on Server-Side Cursors

Server-side cursors are implemented beginning with the C API in MySQL 5.0.2 via the mysql_stmt_attr_set() function. A server-side cursor allows a result set to be generated on the server side, but not transferred to the client except for those rows that the client requests. For example, if a client executes a query but is only interested in the first row, the remaining rows are not transferred.

In MySQL, a server-side cursor is materialized into a temporary table. Initially, this is a MEMORY table, but is converted to a MyISAM table if its size reaches the value of the max_heap_table_size system variable. (Beginning with MySQL 5.0.14, the same temporary-table implementation also is used for cursors in stored routines.) One limitation of the implementation is that for a large result set, retrieving its rows through a cursor might be slow.

Cursors are read only; you cannot use a cursor to update rows.

UPDATE WHERE CURRENT OF and DELETE WHERE CURRENT OF are not implemented, because updatable cursors are not supported.

Cursors are non-holdable (not held open after a commit).

Cursors are asensitive.

Cursors are non-scrollable.

Cursors are not named. The statement handler acts as the cursor ID.

You can have open only a single cursor per prepared statement. If you need several cursors, you must prepare several statements.

You cannot use a cursor for a statement that generates a result set if the statement is not supported in prepared mode. This includes statements such as CHECK TABLES, HANDLER READ, and SHOW BINLOG EVENTS.

F.3. Restrictions on Subqueries

  • In MySQL 5.0 before 5.0.36, if you compare a NULL value to a subquery using ALL, ANY, or SOME, and the subquery returns an empty result, the comparison might evaluate to the non-standard result of NULL rather than to TRUE or FALSE.

  • A subquery's outer statement can be any one of: SELECT, INSERT, UPDATE, DELETE, SET, or DO.

  • Subquery optimization for IN is not as effective as for the = operator or for the IN(value_list) operator.

    A typical case for poor IN subquery performance is when the subquery returns a small number of rows but the outer query returns a large number of rows to be compared to the subquery result.

    The problem is that, for a statement that uses an IN subquery, the optimizer rewrites it as a correlated subquery. Consider the following statement that uses an uncorrelated subquery:

    SELECT ... FROM t1 WHERE t1.a IN (SELECT b FROM t2);
    

    The optimizer rewrites the statement to a correlated subquery:

    SELECT ... FROM t1 WHERE EXISTS (SELECT 1 FROM t2 WHERE t2.b = t1.a);
    

    If the inner and outer queries return M and N rows, respectively, the execution time becomes on the order of O(M×N), rather than O(M+N) as it would be for an uncorrelated subquery.

    An implication is that an IN subquery can be much slower than a query written using an IN(value_list) operator that lists the same values that the subquery would return.

  • In general, you cannot modify a table and select from the same table in a subquery. For example, this limitation applies to statements of the following forms:

    DELETE FROM t WHERE ... (SELECT ... FROM t ...);
    UPDATE t ... WHERE col = (SELECT ... FROM t ...);
    {INSERT|REPLACE} INTO t (SELECT ... FROM t ...);
    

    Exception: The preceding prohibition does not apply if you are using a subquery for the modified table in the FROM clause. Example:

    UPDATE t ... WHERE col = (SELECT (SELECT ... FROM t...) AS _t ...);
    

    Here the prohibition does not apply because the result from a subquery in the FROM clause is stored as a temporary table, so the relevant rows in t have already been selected by the time the update to t takes place.

  • Row comparison operations are only partially supported:

    • For expr IN (subquery), expr can be an n-tuple (specified via row constructor syntax) and the subquery can return rows of n-tuples.

    • For expr op {ALL|ANY|SOME} (subquery), expr must be a scalar value and the subquery must be a column subquery; it cannot return multiple-column rows.

    In other words, for a subquery that returns rows of n-tuples, this is supported:

    (val_1, ..., val_n) IN (subquery)
    

    But this is not supported:

    (val_1, ..., val_n) op {ALL|ANY|SOME} (subquery)
    

    The reason for supporting row comparisons for IN but not for the others is that IN is implemented by rewriting it as a sequence of = comparisons and AND operations. This approach cannot be used for ALL, ANY, or SOME.

  • Row constructors are not well optimized. The following two expressions are equivalent, but only the second can be optimized:

    (col1, col2, ...) = (val1, val2, ...)
    col1 = val1 AND col2 = val2 AND ...
    
  • Subqueries in the FROM clause cannot be correlated subqueries. They are materialized (executed to produce a result set) before evaluating the outer query, so they cannot be evaluated per row of the outer query.

  • The optimizer is more mature for joins than for subqueries, so in many cases a statement that uses a subquery can be executed more efficiently if you rewrite it as a join.

    An exception occurs for the case where an IN subquery can be rewritten as a SELECT DISTINCT join. Example:

    SELECT col FROM t1 WHERE id_col IN (SELECT id_col2 FROM t2 WHERE condition);
    

    That statement can be rewritten as follows:

    SELECT DISTINCT col FROM t1, t2 WHERE t1.id_col = t2.id_col AND condition;
    

    But in this case, the join requires an extra DISTINCT operation and is not more efficient than the subquery.

  • Possible future optimization: MySQL does not rewrite the join order for subquery evaluation. In some cases, a subquery could be executed more efficiently if MySQL rewrote it as a join. This would give the optimizer a chance to choose between more execution plans. For example, it could decide whether to read one table or the other first.

    Example:

    SELECT a FROM outer_table AS ot
    WHERE a IN (SELECT a FROM inner_table AS it WHERE ot.b = it.b);
    

    For that query, MySQL always scans outer_table first and then executes the subquery on inner_table for each row. If outer_table has a lot of rows and inner_table has few rows, the query probably will not be as fast as it could be.

    The preceding query could be rewritten like this:

    SELECT a FROM outer_table AS ot, inner_table AS it
    WHERE ot.a = it.a AND ot.b = it.b;
    

    In this case, we can scan the small table (inner_table) and look up rows in outer_table, which will be fast if there is an index on (ot.a,ot.b).

  • Possible future optimization: A correlated subquery is evaluated for each row of the outer query. A better approach is that if the outer row values do not change from the previous row, do not evaluate the subquery again. Instead, use its previous result.

  • Possible future optimization: A subquery in the FROM clause is evaluated by materializing the result into a temporary table, and this table does not use indexes. This does not allow the use of indexes in comparison with other tables in the query, although that might be useful.

  • Possible future optimization: If a subquery in the FROM clause resembles a view to which the merge algorithm can be applied, rewrite the query and apply the merge algorithm so that indexes can be used. The following statement contains such a subquery:

    SELECT * FROM (SELECT * FROM t1 WHERE t1.t1_col) AS _t1, t2 WHERE t2.t2_col;
    

    The statement can be rewritten as a join like this:

    SELECT * FROM t1, t2 WHERE t1.t1_col AND t2.t2_col;
    

    This type of rewriting would provide two benefits:

    • It avoids the use of a temporary table for which no indexes can be used. In the rewritten query, the optimizer can use indexes on t1.

    • It gives the optimizer more freedom to choose between different execution plans. For example, rewriting the query as a join allows the optimizer to use t1 or t2 first.

  • Possible future optimization: For IN, = ANY, <> ANY, = ALL, and <> ALL with uncorrelated subqueries, use an in-memory hash for a result or a temporary table with an index for larger results. Example:

    SELECT a FROM big_table AS bt
    WHERE non_key_field IN (SELECT non_key_field FROM table WHERE condition)
    

    In this case, we could create a temporary table:

    CREATE TABLE t (key (non_key_field))
    (SELECT non_key_field FROM table WHERE condition)
    

    Then, for each row in big_table, do a key lookup in t based on bt.non_key_field.

F.4. Restrictions on Views

View processing is not optimized:

  • It is not possible to create an index on a view.

  • Indexes can be used for views processed using the merge algorithm. However, a view that is processed with the temptable algorithm is unable to take advantage of indexes on its underlying tables (although indexes can be used during generation of the temporary tables).

Subqueries cannot be used in the FROM clause of a view. This limitation will be lifted in the future.

There is a general principle that you cannot modify a table and select from the same table in a subquery. See Section F.3, “Restrictions on Subqueries”.

The same principle also applies if you select from a view that selects from the table, if the view selects from the table in a subquery and the view is evaluated using the merge algorithm. Example:

CREATE VIEW v1 AS
SELECT * FROM t2 WHERE EXISTS (SELECT 1 FROM t1 WHERE t1.a = t2.a);

UPDATE t1, v2 SET t1.a = 1 WHERE t1.b = v2.b;

If the view is evaluated using a temporary table, you can select from the table in the view subquery and still modify that table in the outer query. In this case the view will be stored in a temporary table and thus you are not really selecting from the table in a subquery and modifying it “at the same time.” (This is another reason you might wish to force MySQL to use the temptable algorithm by specifying ALGORITHM = TEMPTABLE in the view definition.)

You can use DROP TABLE or ALTER TABLE to drop or alter a table that is used in a view definition (which invalidates the view) and no warning results from the drop or alter operation. An error occurs later when the view is used.

A view definition is “frozen” by certain statements:

  • If a statement prepared by PREPARE refers to a view, the view contents seen each time the statement is executed later will be the contents of the view at the time it was prepared. This is true even if the view definition is changed after the statement is prepared and before it is executed. Example:

    CREATE VIEW v AS SELECT 1;
    PREPARE s FROM 'SELECT * FROM v';
    ALTER VIEW v AS SELECT 2;
    EXECUTE s;
    

    The result returned by the EXECUTE statement is 1, not 2.

  • If a statement in a stored routine refers to a view, the view contents seen by the statement are its contents the first time that statement is executed. For example, this means that if the statement is executed in a loop, further iterations of the statement see the same view contents, even if the view definition is changed later in the loop. Example:

    CREATE VIEW v AS SELECT 1;
    delimiter //
    CREATE PROCEDURE p ()
    BEGIN
      DECLARE i INT DEFAULT 0;
      WHILE i < 5 DO
        SELECT * FROM v;
        SET i = i + 1;
        ALTER VIEW v AS SELECT 2;
      END WHILE;
    END;
    //
    delimiter ;
    CALL p();
    

    When the procedure p() is called, the SELECT returns 1 each time through the loop, even though the view definition is changed within the loop.

With regard to view updatability, the overall goal for views is that if any view is theoretically updatable, it should be updatable in practice. This includes views that have UNION in their definition. Currently, not all views that are theoretically updatable can be updated. The initial view implementation was deliberately written this way to get usable, updatable views into MySQL as quickly as possible. Many theoretically updatable views can be updated now, but limitations still exist:

  • Updatable views with subqueries anywhere other than in the WHERE clause. Some views that have subqueries in the SELECT list may be updatable.

  • You cannot use UPDATE to update more than one underlying table of a view that is defined as a join.

  • You cannot use DELETE to update a view that is defined as a join.

There exists a shortcoming with the current implementation of views. If a user is granted the basic privileges necessary to create a view (the CREATE VIEW and SELECT privileges), that user will be unable to call SHOW CREATE VIEW on that object unless the user is also granted the SHOW VIEW privilege.

That shortcoming can lead to problems backing up a database with mysqldump, which may fail due to insufficient privileges. This problem is described in Bug#22062.

The workaround to the problem is for the administrator to manually grant the SHOW VIEW privilege to users who are granted CREATE VIEW, since MySQL doesn't grant it implicitly when views are created.

Views do not have indexes, so index hints do not apply. Use of index hints when selecting from a view is disallowed.

F.5. Restrictions on XA Transactions

XA transaction support is limited to the InnoDB storage engine.

The MySQL XA implementation is for “external XA,” where a MySQL server acts as a Resource Manager and client programs act as Transaction Managers. “Internal XA” is not implemented. This would allow individual storage engines within a MySQL server to act as RMs, and the server itself to act as a TM. Internal XA is required for handling XA transactions that involve more than one storage engine. The implementation of internal XA is incomplete because it requires that a storage engine support two-phase commit at the table handler level, and currently this is true only for InnoDB.

For XA START, the JOIN and RESUME clauses are not supported.

For XA END, the SUSPEND [FOR MIGRATE] clause is not supported.

The requirement that the bqual part of the xid value be different for each XA transaction within a global transaction is a limitation of the current MySQL XA implementation. It is not part of the XA specification.

If an XA transaction has reached the PREPARED state and the MySQL server is killed (for example, with kill -9 on Unix) or shuts down abnormally, the transaction can be continued after the server restarts. However, if the client reconnects and commits the transaction, the transaction will be absent from the binary log even though it has been committed. This means the data and the binary log have gone out of synchrony. An implication is that XA cannot be used safely together with replication.

It is possible that the server will roll back a pending XA transaction, even one that has reached the PREPARED state. This happens if a client connection terminates and the server continues to run, or if clients are connected and the server shuts down gracefully. (In the latter case, the server marks each connection to be terminated, and then rolls back the PREPARED XA transaction associated with it.) It should be possible to commit or roll back a PREPARED XA transaction, but this cannot be done without changes to the binary logging mechanism.

F.6. Restrictions on Character Sets

  • Identifiers are stored in mysql database tables (user, db, and so forth) using utf8, but identifiers can contain only characters in the Basic Multilingual Plane (BMP). Supplementary characters are not allowed in identifiers.

  • The ucs2 character sets has the following restrictions:

    • It cannot be used as a client character set, which means that it does not work for SET NAMES or SET CHARACTER SET. (See Section 9.1.4, “Connection Character Sets and Collations”.)

    • It is currently not possible to use LOAD DATA INFILE to load data files that use this character set.

    • FULLTEXT indexes cannot be created on a column that this character set. However, you can perform IN BOOLEAN MODE searches on the column without an index.

  • The REGEXP and RLIKE operators work in byte-wise fashion, so they are not multi-byte safe and may produce unexpected results with multi-byte character sets. In addition, these operators compare characters by their byte values and accented characters may not compare as equal even if a given collation treats them as equal.

F.7. Limits in MySQL

This section lists current limits in MySQL 5.0.

F.7.1. Limits of Joins

The maximum number of tables that can be referenced in a single join is 61. This also applies to the number of tables that can be referenced in the definition of a view.

F.7.2. The Maximum Number of Columns Per Table

There is a hard limit of 4096 columns per table, but the effective maximum may be less for a given table. The exact limit depends on several interacting factors, listed in the following discussion.

  • Every table has a maximum row size of 65,535 bytes. This maximum applies to all storage engines, but a given engine might have additional constraints that result in a lower effective maximum row size.

    The maximum row size constrains the number of columns because the total width of all columns cannot exceed this size. For example, utf8 characters require up to three bytes per character, so for a CHAR(255) CHARACTER SET utf8 column, the server must allocate 255 × 3 = 765 bytes per value. Consequently, a table cannot contain more than 65,535 / 765 = 85 such columns.

    Storage for variable-length columns includes length bytes, which are assessed against the row size. For example, a VARCHAR(255) CHARACTER SET utf8 column takes two bytes to store the length of the value, so each value can take up to 767 bytes.

    BLOB and TEXT columns count from one to four plus eight bytes each toward the row-size limit because their contents are stored separately.

    Declaring columns NULL can reduce the maximum number of columns allowed. NULL columns require additional space in the row to record whether or not their values are NULL.

    For MyISAM tables, each NULL column takes one bit extra, rounded up to the nearest byte. The maximum row length in bytes can be calculated as follows:

    row length = 1
                 + (sum of column lengths)
                 + (number of NULL columns + delete_flag + 7)/8
                 + (number of variable-length columns)
    

    delete_flag is 1 for tables with static row format. Static tables use a bit in the row record for a flag that indicates whether the row has been deleted. delete_flag is 0 for dynamic tables because the flag is stored in the dynamic row header.

    These calculations do not apply for InnoDB tables, for which storage size is no different for NULL columns than for NOT NULL columns.

    The following statement to create table t1 succeeds because the columns require 32,765 + 2 bytes and 32,766 + 2 bytes, which falls within the maximum row size of 65,535 bytes:

    mysql> CREATE TABLE t1
        -> (c1 VARCHAR(32765) NOT NULL, c2 VARCHAR(32766) NOT NULL);
    Query OK, 0 rows affected (0.01 sec)
    

    The following statement to create table t2 fails because the columns are NULL and require additional space that causes the row size to exceed 65,535 bytes:

    mysql> CREATE TABLE t2
        -> (c1 VARCHAR(32765) NULL, c2 VARCHAR(32766) NULL);
    ERROR 1118 (42000): Row size too large. The maximum row size for the
    used table type, not counting BLOBs, is 65535. You have to change some
    columns to TEXT or BLOBs
    
  • Each table has an .frm file that contains the table definition. The .frm file size limit is fixed at 64KB. If a table definition reaches this size, no more columns can be added. The expression that checks information to be stored in the .frm file against the limit looks like this:

    if (info_length+(ulong) create_fields.elements*FCOMP+288+
        n_length+int_length+com_length > 65535L || int_count > 255)
    

    The relevant factors in this expression are:

    • info_length is space needed for “screens.” This is related to MySQL's Unireg heritage.

    • create_fields.elements is the number of columns.

    • FCOMP is 17.

    • n_length is the total length of all column names, including one byte per name as a separator.

    • int_length is related to the list of values for SET and ENUM columns.

    • com_length is the total length of column and table comments.

    Thus, using long column names can reduce the maximum number of columns, as can the inclusion of ENUM or SET columns, or use of column or table comments.

  • Individual storage engines might impose additional restrictions that limit table column count. Examples:

    • InnoDB allows no more than 1000 columns.

    • InnoDB restricts row size to something less than half a database page (approximately 8000 bytes), not including VARBINARY, VARCHAR, BLOB, or TEXT columns.

    • Different InnoDB storage formats (COMPRESSED, REDUNDANT) use different amounts of page header and trailer data, which affects the amount of storage available for rows.

F.7.3. Windows Platform Limitations

The following limitations apply only to the Windows platform:

  • The number of open file descriptors on Windows is limited to a maximum of 2048, which may limit the ability to open a large number of tables simultaneously. This limit is due to the compatibility functions used to open files on Windows that use the POSIX compatibility layer.

    This limitation will also cause problems if you try to set max_open_files to a value greater than the 2048 file limit.

  • On Windows 32-bit platforms it is not possible to use more than 2GB of RAM within a single process, including MySQL. This is because the physical address limit on Windows 32-bit is 4GB and the default setting within Windows is to split the virtual address space between kernel (2GB) and user/applications (2GB).

    To use more memory than this you will need to use a 64-bit version of Windows.

  • When using MyISAM tables, you cannot use aliases within Windows link to the data files on another volume and then link back to the main MySQL datadir location.

    This facility is often used to move the data and index files to a RAID or other fast solution, while retaining the main .FRM files in the default data directory configured with the datadir option.

  • The timers within MySQL used on Windows are of a lower precision than the timers used on Linux. For most situations you may not notice a difference, but the delay implied by a call to SLEEP() on Windows and Linux may differ slightly due to the differences in precision.

  • There is no 64-bit OLEDB Provider for ODBC (MSDASQL) in any 64-bit Windows operating system up to and including Windows Vista. In practical terms this means that you can't use the MySQL ODBC driver from ADO and other users of OLEDB.