11.8 Data Type Storage Requirements

See Section 15.10, “InnoDB Row Formats” for information about storage requirements for InnoDB tables.

Every table using the NDB storage engine requires a primary key; if you do not define a primary key, a “hidden” primary key is created by NDB. This hidden primary key consumes 31-35 bytes per table record.

You can use the ndb_size.pl Perl script to estimate NDB storage requirements. It connects to a current MySQL (not NDB Cluster) database and creates a report on how much space that database would require if it used the NDB storage engine. See Section 22.4.28, “ndb_size.pl — NDBCLUSTER Size Requirement Estimator” for more information.

Values for DECIMAL (and NUMERIC) columns are represented using a binary format that packs nine decimal (base 10) digits into four bytes. Storage for the integer and fractional parts of each value are determined separately. Each multiple of nine digits requires four bytes, and the “leftover” digits require some fraction of four bytes. The storage required for excess digits is given by the following table.

For TIME, DATETIME, and TIMESTAMP columns, the storage required for tables created before MySQL 5.6.4 differs from tables created from 5.6.4 on. This is due to a change in 5.6.4 that permits these types to have a fractional part, which requires from 0 to 3 bytes.

As of MySQL 5.6.4, storage for YEAR and DATE remains unchanged. However, TIME, DATETIME, and TIMESTAMP are represented differently. DATETIME is packed more efficiently, requiring 5 rather than 8 bytes for the nonfractional part, and all three parts have a fractional part that requires from 0 to 3 bytes, depending on the fractional seconds precision of stored values.

For example, TIME(0), TIME(2), TIME(4), and TIME(6) use 3, 4, 5, and 6 bytes, respectively. TIME and TIME(0) are equivalent and require the same storage.

For details about internal representation of temporal values, see MySQL Internals: Important Algorithms and Structures.

In the following table, M represents the declared column length in characters for nonbinary string types and bytes for binary string types. L represents the actual length in bytes of a given string value.

Variable-length string types are stored using a length prefix plus data. The length prefix requires from one to four bytes depending on the data type, and the value of the prefix is L (the byte length of the string). For example, storage for a MEDIUMTEXT value requires L bytes to store the value plus three bytes to store the length of the value.

To calculate the number of bytes used to store a particular CHAR, VARCHAR, or TEXT column value, you must take into account the character set used for that column and whether the value contains multibyte characters. In particular, when using a utf8 Unicode character set, you must keep in mind that not all characters use the same number of bytes. utf8mb3 and utf8mb4 character sets can require up to three and four bytes per character, respectively. For a breakdown of the storage used for different categories of utf8mb3 or utf8mb4 characters, see Section 10.9, “Unicode Support”.

VARCHAR, VARBINARY, and the BLOB and TEXT types are variable-length types. For each, the storage requirements depend on these factors:

For example, a VARCHAR(255) column can hold a string with a maximum length of 255 characters. Assuming that the column uses the latin1 character set (one byte per character), the actual storage required is the length of the string (L), plus one byte to record the length of the string. For the string 'abcd', L is 4 and the storage requirement is five bytes. If the same column is instead declared to use the ucs2 double-byte character set, the storage requirement is 10 bytes: The length of 'abcd' is eight bytes and the column requires two bytes to store lengths because the maximum length is greater than 255 (up to 510 bytes).

The effective maximum number of bytes that can be stored in a VARCHAR or VARBINARY column is subject to the maximum row size of 65,535 bytes, which is shared among all columns. For a VARCHAR column that stores multibyte characters, the effective maximum number of characters is less. For example, utf8mb4 characters can require up to four bytes per character, so a VARCHAR column that uses the utf8mb4 character set can be declared to be a maximum of 16,383 characters. See Section C.10.4, “Limits on Table Column Count and Row Size”.

InnoDB encodes fixed-length fields greater than or equal to 768 bytes in length as variable-length fields, which can be stored off-page. For example, a CHAR(255) column can exceed 768 bytes if the maximum byte length of the character set is greater than 3, as it is with utf8mb4.

The NDB storage engine supports variable-width columns. This means that a VARCHAR column in an NDB Cluster table requires the same amount of storage as would any other storage engine, with the exception that such values are 4-byte aligned. Thus, the string 'abcd' stored in a VARCHAR(50) column using the latin1 character set requires 8 bytes (rather than 5 bytes for the same column value in a MyISAM table).

TEXT and BLOB columns are implemented differently in NDB; each row in a TEXT column is made up of two separate parts. One of these is of fixed size (256 bytes), and is actually stored in the original table. The other consists of any data in excess of 256 bytes, which is stored in a hidden table. The rows in this second table are always 2000 bytes long. This means that the size of a TEXT column is 256 if size <= 256 (where size represents the size of the row); otherwise, the size is 256 + size + (2000 × (size − 256) % 2000).

The size of an ENUM object is determined by the number of different enumeration values. One byte is used for enumerations with up to 255 possible values. Two bytes are used for enumerations having between 256 and 65,535 possible values. See Section 11.4.4, “The ENUM Type”.

The size of a SET object is determined by the number of different set members. If the set size is N, the object occupies (N+7)/8 bytes, rounded up to 1, 2, 3, 4, or 8 bytes. A SET can have a maximum of 64 members. See Section 11.4.5, “The SET Type”.

MySQL stores geometry values using 4 bytes to indicate the SRID followed by the WKB representation of the value. The LENGTH() function returns the space in bytes required for value storage.

For descriptions of WKB and internal storage formats for spatial values, see Section 11.5.3, “Supported Spatial Data Formats”.

In general, the storage requirement for a JSON column is approximately the same as for a LONGBLOB or LONGTEXT column; that is, the space consumed by a JSON document is roughly the same as it would be for the document's string representation stored in a column of one of these types. However, there is an overhead imposed by the binary encoding, including metadata and dictionaries needed for lookup, of the individual values stored in the JSON document. For example, a string stored in a JSON document requires 4 to 10 bytes additional storage, depending on the length of the string and the size of the object or array in which it is stored.

In addition, MySQL imposes a limit on the size of any JSON document stored in a JSON column such that it cannot be any larger than the value of max_allowed_packet.