Loops that need transactions normally can be coded with the help of LOCK TABLES, and you don't need cursors to update records on the fly.

To avoid using ROLLBACK, you can use the following strategy:

This is usually a much faster method than using transactions with possible rollbacks, although not always. The only situation this solution doesn't handle is when someone kills the threads in the middle of an update. In this case, all locks will be released but some of the updates may not have been executed.

You can also use functions to update records in a single operation. You can get a very efficient application by using the following techniques:

For example, when we are doing updates to some customer information, we update only the customer data that has changed and test only that none of the changed data, or data that depends on the changed data, has changed compared to the original row. The test for changed data is done with the WHERE clause in the UPDATE statement. If the record wasn't updated, we give the client a message: “Some of the data you have changed has been changed by another user.” Then we show the old row versus the new row in a window so that the user can decide which version of the customer record to use.

This gives us something that is similar to column locking but is actually even better because we only update some of the columns, using values that are relative to their current values. This means that typical UPDATE statements look something like these:

UPDATE tablename SET pay_back=pay_back+125;

UPDATE customer
  SET
    customer_date='current_date',
    address='new address',
    phone='new phone',
    money_owed_to_us=money_owed_to_us-125
  WHERE
    customer_id=id AND address='old address' AND phone='old phone';

This is very efficient and works even if another client has changed the values in the pay_back or money_owed_to_us columns.

In many cases, users have wanted LOCK TABLES and/or ROLLBACK for the purpose of managing unique identifiers. This can be handled much more efficiently without locking or rolling back by using an AUTO_INCREMENT column and either the LAST_INSERT_ID() SQL function or the mysql_insert_id() C API function. See the section called “Information Functions”. See mysql_insert_id().

You can generally code around the need for row-level locking. Some situations really do need it, and InnoDB tables support row-level locking. With MyISAM tables, you can use a flag column in the table and do something like the following:

UPDATE tbl_name SET row_flag=1 WHERE id=ID;

MySQL returns 1 for the number of affected rows if the row was found and row_flag wasn't already 1 in the original row.

You can think of it as though MySQL Server changed the preceding query to:

UPDATE tbl_name SET row_flag=1 WHERE id=ID AND row_flag <> 1;

Assuming proper design of the relationships, foreign key constraints make it more difficult for a programmer to introduce an inconsistency into the database.

Centralized checking of constraints by the database server makes it unnecessary to perform these checks on the application side. This eliminates the possibility that different applications may not all check the constraints in the same way.

Using cascading updates and deletes can simplify the application code.

Properly designed foreign key rules aid in documenting relationships between tables.

In the absence of server-side foreign key relationship checking, the application itself must handle relationship issues. For example, it must take care to insert rows into tables in the proper order, and to avoid creating orphaned child records. It must also be able to recover from errors that occur in the middle of multiple-record insert operations.

If ON DELETE is the only referential integrity capability an application needs, note that as of MySQL Server 4.0, you can use multiple-table DELETE statements to delete rows from many tables with a single statement. See DELETE.

A workaround for the lack of ON DELETE is to add the appropriate DELETE statement to your application when you delete records from a table that has a foreign key. In practice, this is often as quick as using foreign keys, and is more portable.

Foreign key support addresses many referential integrity issues, but it is still necessary to design key relationships carefully to avoid circular rules or incorrect combinations of cascading deletes.

It is not uncommon for a DBA to create a topology of relationships that makes it difficult to restore individual tables from a backup. (MySQL alleviates this difficulty by allowing you to temporarily disable foreign key checks when reloading a table that depends on other tables. See the section called “FOREIGN KEY Constraints”. As of MySQL 4.1.1, mysqldump generates dump files that take advantage of this capability automatically when reloaded.)

If you try to store a value outside the range in a numerical column, MySQL Server instead stores zero, the smallest possible value, or the largest possible value in the column.

For strings, MySQL stores either the empty string or the longest possible string that can be in the column.

If you try to store a string that doesn't start with a number into a numerical column, MySQL Server stores 0.

If you try to store NULL into a column that doesn't take NULL values, MySQL Server stores 0 or '' (the empty string) instead. This last behavior can, for single-row inserts, be changed when MySQL is built by using the -DDONT_USE_DEFAULT_FIELDS compile option.) See configure options. This causes INSERT statements to generate an error unless you explicitly specify values for all columns that require a non-NULL value.

MySQL allows you to store some incorrect date values into DATE and DATETIME columns (like '2000-02-31' or '2000-02-00'). The idea is that it's not the job of the SQL server to validate dates. If MySQL can store a date value and retrieve exactly the same value, MySQL stores it as given. If the date is totally wrong (outside the server's ability to store it), the special date value '0000-00-00' is stored in the column instead.

If you don't specify a column that doesn't have a default value in an INSERT statement, MySQL will insert an empty string for string columns and 0 for numerical columns.

An ENUM value must be one of those listed in the column definition, or the internal numeric equivalent thereof. The value cannot be the error value (that is, 0 or the empty string). For a column defined as ENUM('a','b','c'), values such as '', 'd', and 'ax' are illegal.

A SET value must be the empty string or a value consisting of one or more of the values listed in the column definition separated by commas. For a column defined as SET('a','b','c'), values such as 'd', and 'a,b,c,d' are illegal.

You should avoid using spaces at the end of column names because this can cause weird behavior. (Fixed in MySQL 4.0.) (Bug #4196)

You can get a deadlock (hung thread) if you use LOCK TABLE to lock multiple tables and then in the same connection use DROP TABLE to drop one of them while another thread is trying to lock it. (To break the deadlock, you can use KILL to terminate any of the threads involved.) This issue is resolved as of MySQL 4.0.12.

SELECT MAX(key_column) FROM t1,t2,t3... where one of the tables are empty doesn't return NULL but instead returns the maximum value for the column. This issue is resolved as of MySQL 4.0.11.

DELETE FROM heap_table without a WHERE clause doesn't work on a locked HEAP table.

In a UNION, the first SELECT determines the type, max_length, and NULL properties for the resulting columns. This issue is resolved as of MySQL 4.1.1; the property values are based on the rows from all UNION parts.

In DELETE with many tables, you can't refer to tables to be deleted through an alias. This is fixed as of MySQL 4.1.

You cannot mix UNION ALL and UNION DISTINCT in the same query. If you use ALL for one UNION, it is used for all of them. This is fixed as of MySQL 4.1.2. The rules for mixed UNION types are given in UNION.

FLUSH TABLES WITH READ LOCK does not block CREATE TABLE, which may cause a problem with the binary log position when doing a full backup of tables and the binary log.

Even if you are using lower_case_table_names=2 (which enables MySQL to remember the used case for databases and table names) MySQL will not on case insensitive systems remember the used case for database names for the function DATABASE() or in various logs.

Dropping a FOREIGN KEY constraint doesn't work in replication because the constraint may have another name on the slave.

REPLACE (and LOAD DATA with the REPLACE option) does not trigger ON DELETE CASCADE.

DISTINCT with ORDER BY doesn't work inside GROUP_CONCAT() if you don't use all and only those columns that are in the DISTINCT list.

If one user has a long-running transaction and another user drops a table that is updated in the transaction, there is small chance that the binary log may contain the DROP TABLE command before the table is used in the transaction itself. We plan to fix this in 5.0 by having the DROP TABLE wait until the table is not used in any transaction.

When inserting a big integer value (between 2^63 and 2^64−1) into a decimal/string column, it is inserted as a negative value because the number is evaluated in a signed integer context. We plan to fix this in MySQL 4.1.

FLUSH TABLES WITH READ LOCK does not block COMMIT if the server is running without binary logging, which may cause a problem (of consistency between tables) when doing a full backup.

ANALYZE TABLE on a BDB table may in some cases make the table unusable until you restart mysqld. If this happens, you will see errors of the following form in the MySQL error file:

001207 22:07:56  bdb:  log_flush: LSN past current end-of-log

MySQL accepts parentheses in the FROM clause of a SELECT statement, but silently ignores them. The reason for not giving an error is that many clients that automatically generate queries add parentheses in the FROM clause even where they are not needed.

Concatenating many RIGHT JOINS or combining LEFT and RIGHT join in the same query may not give a correct answer because MySQL only generates NULL rows for the table preceding a LEFT or before a RIGHT join. This will be fixed in 5.0 at the same time we add support for parentheses in the FROM clause.

Don't execute ALTER TABLE on a BDB table on which you are running multiple-statement transactions until all those transactions complete. (The transaction will probably be ignored.)

ANALYZE TABLE, OPTIMIZE TABLE, and REPAIR TABLE may cause problems on tables for which you are using INSERT DELAYED.

Doing a LOCK TABLE ... and FLUSH TABLES ... doesn't guarantee that there isn't a half-finished transaction in progress on the table.

BDB tables are a bit slow to open. If you have many BDB tables in a database, it will take a long time to use the mysql client on the database if you are not using the -A option or if you are using rehash. This is especially notable when you have a large table cache.

Replication uses query-level logging: The master writes the executed queries to the binary log. This is a very fast, compact, and efficient logging method that works perfectly in most cases. Although we have never heard of it actually occurring, it is theoretically possible for the data on the master and slave to become different if a query is designed in such a way that the data modification is non-deterministic; that is, left to the will of the query optimizer. (That generally is not a good practice anyway, even outside of replication!) For example:

If and only if all these queries have no ORDER BY clause guaranteeing a deterministic order.

For example, for INSERT ... SELECT with no ORDER BY, the SELECT may return rows in a different order (which will result in a row having different ranks, hence getting a different number in the AUTO_INCREMENT column), depending on the choices made by the optimizers on the master and slave. A query will be optimized differently on the master and slave only if:

This problem may also affect database restoration using mysqlbinlog|mysql.

The easiest way to avoid this problem in all cases is to add an ORDER BY clause to such non-deterministic queries to ensure that the rows are always stored or modified in the same order. In future MySQL versions, we will automatically add an ORDER BY clause when needed.

Log filenames are based on the server hostname (if you don't specify a filename with the startup option). For now you have to use options like --log-bin=old_host_name-bin if you change your hostname to something else. Another option is to just rename the old files to reflect your hostname change. See the section called “mysqld Command-Line Options”.

mysqlbinlog will not delete temporary files left after a LOAD DATA INFILE command. See mysqlbinlog.

RENAME doesn't work with TEMPORARY tables or tables used in a MERGE table.

When using the RPAD() function in a query that has to be resolved by using a temporary table, all resulting strings will have rightmost spaces removed. This is an example of such a query:

SELECT RPAD(t1.column1, 50, ' ') AS f2, RPAD(t2.column2, 50, ' ') AS f1
FROM table1 as t1 LEFT JOIN table2 AS t2 ON t1.record=t2.joinID
ORDER BY t2.record;

The final result of this bug is that you will not be able to get spaces on the right side of the resulting values. The problem also occurs for any other string function that adds spaces to the right.

The reason for this is due to the fact that HEAP tables, which are used first for temporary tables, are not capable of handling VARCHAR columns.

This behavior exists in all versions of MySQL. It will be fixed in one of the 4.1 series releases.

Due to the way table definition files are stored, you cannot use character 255 (CHAR(255)) in table names, column names, or enumerations. This is scheduled to be fixed in version 5.1 when we have new table definition format files.

When using SET CHARACTER SET, you can't use translated characters in database, table, and column names.

You can't use ‘_’ or ‘%’ with ESCAPE in LIKE ... ESCAPE.

If you have a DECIMAL column in which the same number is stored in different formats (for example, +01.00, 1.00, 01.00), GROUP BY may regard each value as a different value.

You cannot build the server in another directory when using MIT-pthreads. Because this requires changes to MIT-pthreads, we are not likely to fix this. See the section called “MIT-pthreads Notes”.

BLOB and TEXTvalues can't “reliably” be used in GROUP BY or ORDER BY or DISTINCT. Only the first max_sort_length bytes are used when comparing BLOB values in these cases. The default value of max_sort_length value is 1024. It can be changed at server startup time. As of MySQL 4.0.3, it can also be changed at runtime. For older versions, a workaround for most cases is to use a substring. For example:

SELECT DISTINCT LEFT(blob_col,2048) FROM tbl_name;

Numeric calculations are done with BIGINT or DOUBLE (both are normally 64 bits long). Which precision you get depends on the function. The general rule is that bit functions are done with BIGINT precision, IF and ELT() with BIGINT or DOUBLE precision, and the rest with DOUBLE precision. You should try to avoid using unsigned long long values if they resolve to be bigger than 63 bits (9223372036854775807) for anything other than bit fields. MySQL Server 4.0 has better BIGINT handling than 3.23.

All string columns, except BLOB and TEXT columns, automatically have all trailing spaces removed when retrieved. For CHAR types, this is okay. The bug is that in MySQL Server, VARCHAR columns are treated the same way.

You can have only up to 255 ENUM and SET columns in one table.

In MIN(), MAX(), and other aggregate functions, MySQL currently compares ENUM and SET columns by their string value rather than by the string's relative position in the set.

mysqld_safe redirects all messages from mysqld to the mysqld log. One problem with this is that if you execute mysqladmin refresh to close and reopen the log, stdout and stderr are still redirected to the old log. If you use --log extensively, you should edit mysqld_safe to log to host_name.err instead of host_name.log so that you can easily reclaim the space for the old log by deleting the old one and executing mysqladmin refresh.

In the UPDATE statement, columns are updated from left to right. If you refer to an updated column, you get the updated value instead of the original value. For example, the following statement increments KEY by 2, not 1:

mysql> UPDATE tbl_name SET KEY=KEY+1,KEY=KEY+1;

You can refer to multiple temporary tables in the same query, but you cannot refer to any given temporary table more than once. For example, the following doesn't work:

mysql> SELECT * FROM temp_table, temp_table AS t2;
ERROR 1137: Can't reopen table: 'temp_table'

The optimizer may handle DISTINCT differently when you are using “hidden” columns in a join than when you are not. In a join, hidden columns are counted as part of the result (even if they are not shown), whereas in normal queries, hidden columns don't participate in the DISTINCT comparison. We will probably change this in the future to never compare the hidden columns when executing DISTINCT.

An example of this is:

SELECT DISTINCT mp3id FROM band_downloads
       WHERE userid = 9 ORDER BY id DESC;

and

SELECT DISTINCT band_downloads.mp3id
       FROM band_downloads,band_mp3
       WHERE band_downloads.userid = 9
       AND band_mp3.id = band_downloads.mp3id
       ORDER BY band_downloads.id DESC;

In the second case, you might in MySQL Server 3.23.x get two identical rows in the result set (because the values in the hidden id column may differ).

Note that this happens only for queries where you don't have the ORDER BY columns in the result.

Because MySQL Server allows you to work with table types that don't support transactions, and thus can't roll back data, some things behave a little differently in MySQL Server than in other SQL servers. This is just to ensure that MySQL Server never needs to do a rollback for an SQL statement. This may be a little awkward at times because column values must be checked in the application, but this will actually give you a nice speed increase because it allows MySQL Server to do some optimizations that otherwise would be very hard to do.

If you set a column to an incorrect value, MySQL Server will, instead of doing a rollback, store the “best possible value” in the column. For information about how this occurs, see the section called “How MySQL Deals with Constraints”.

If you execute a PROCEDURE on a query that returns an empty set, in some cases the PROCEDURE will not transform the columns.

Creation of a table of type MERGE doesn't check whether the underlying tables are of compatible types.

If you use ALTER TABLE first to add a UNIQUE index to a table used in a MERGE table and then to add a normal index on the MERGE table, the key order will be different for the tables if there was an old key that was not unique in the table. This is because ALTER TABLE puts UNIQUE indexes before normal indexes to be able to detect duplicate keys as early as possible.

In the following case you can get a core dump:

Before MySQL Server 3.23.2, an UPDATE that updated a key with a WHERE on the same key may have failed because the key was used to search for records and the same row may have been found multiple times:

UPDATE tbl_name SET KEY=KEY+1 WHERE KEY > 100;

A workaround is to use:

UPDATE tbl_name SET KEY=KEY+1 WHERE KEY+0 > 100;

This will work because MySQL Server will not use an index on expressions in the WHERE clause.

Before MySQL Server 3.23, all numeric types were treated as fixed-point fields. That means that you had to specify how many decimals a floating-point field should have. All results were returned with the correct number of decimals.