This chapter compares the MySQL database and the Oracle database. It includes the following sections:
This section includes information about security issues with MySQL databases and Oracle databases.
As with Oracle, MySQL users are maintained by the database. MySQL uses a set of grant tables to keep track of users and the privileges that they can have. MySQL uses these grant tables when performing authentication, authorization and access control for users.
Unlike Oracle (when set up to use database authentication) and most other databases that use only the user name and password to authenticate a user, MySQL uses an additional location parameter when authenticating a user. This location parameter is usually the host name, IP address, or a wildcard (“%”). With this additional parameter, MySQL may further restrict a user access to the database to a particular host or hosts in a domain. Moreover, this also allows a different password and set of privileges to be enforced for a user depending on the host from which the connection is made. Thus, user scott, who logs on from abc.com may or may not the same as user scott who logs on from xyz.com.
The MySQL privilege system is a hierarchical system that works through inheritance. Privileges granted at a higher level are implicitly passed down to all lower levels and may be overridden by the same privileges set at lower levels. MySQL allows privileges to be granted at five different levels, in descending order of the scope of the privileges:
Global
Per-host basis
Database-level
Table-specific
Column-specific (single column in a single table
Each level has a corresponding grant table in the database. When performing a privilege check, MySQL checks each of the tables in descending order of the scope of the privileges, and the privileges granted at a lower level take precedence over the same privileges granted at a higher level.
The privileges supported by MySQL are grouped into two types: administrative privileges and per-object privileges. The administrative privileges are global privileges that have server-wide effects and are concerned with the functioning of MySQL. These administrative privileges include the FILE, PROCESS, REPLICATION, SHUTDOWN and SUPER privilege. The per-object privileges affect database objects such tables, columns, indexes, and stored procedures, and can be granted with a different scope. These per-object privileges are named after the SQL queries that trigger their checks.
Unlike in Oracle, there is no concept of role in MySQL. Thus, in order to grant a group of users the same set of privileges, the privileges have to be granted to each user separately. Alternately, though less satisfactory for auditing, users performing tasks as a role may all share a single user account that is designated for the "role" and with the required privileges granted.
The schema contains the definitions of the tables, views, indexes, users, constraints, stored procedures, triggers, and other database-specific objects. Most relational databases work with similar objects.
This section contains the following:
There are many similarities between schema objects in Oracle and MySQL. However, some schema objects differ between these databases. For more information about schema objects, see Oracle SQL Reference.
Table 2-1 shows the differences between Oracle and MySQL.
Table 2-1 Schema Objects in Oracle and MySQL
| Oracle | MySQL |
|---|---|
|
AFTER trigger |
trigger |
|
BEFORE trigger |
trigger |
|
Check constraint |
Check constraint |
|
Column default |
Column default |
|
Database |
Database |
|
Foreign key |
Foreign key |
|
Index |
Index |
|
Package |
N/A |
|
PL/SQL function |
Routine |
|
PL/SQL procedure |
Routine |
|
Primary key |
Primary key |
|
Role |
N/A |
|
Schema |
Schema |
|
Sequence |
AUTO_INCREMENT for a column |
|
Snapshot |
N/A |
|
Synonym |
N/A |
|
Table |
Table |
|
Tablespace |
N/A |
|
Temporary table |
Temporary table |
|
Trigger for each row |
Trigger for each row |
|
Unique key |
Unique key |
|
User |
User |
|
View |
View |
Oracle is case insensitive to object names, and Oracle schema object names are stored as uppercase.
As in Oracle, column, index, stored procedure, and trigger names as well as column aliases in MySQL are case insensitive on all platforms. However, the case sensitivity of database and tables names for MySQL differs from Oracle. In MySQL, databases correspond to directories within the data directory, and tables correspond to one or more files within the database directory. As such, the case sensitivity of the database and table names is determined by the case sensitivity of the underlying operating systems. This means that database and table names are not case-sensitive in Windows and are case-sensitive in most varieties of Unix. However, MySQL allows users to determine how the database and table names are stored on disk and in their use in MySQL through the lower_case_table_names system variable. Table aliases are case-sensitive in releases before MySQL 4.1.1.
Both Oracle and MySQL let you use reserved words as object names by representing the name with a quoted identifier. However, MySQL allows some reserved words such as DATE and TIMESTAMP to be used as unquoted identifier for object names, although this is not allowed in Oracle. SQL Developer appends an underscore (_) to the name of a MySQL object that is an Oracle reserved word.
MySQL and Oracle have some minor differences in their definition of an identifier. In MySQL, an unquoted identifier may begin with a digit, and double quotation marks are allowed in a quoted identifier; however, neither of these is allowed in an Oracle identifier. In MySQL, the quote character is the backtick (`). If the SQL mode ANSI_QUOTES is set, double quotes can also be used to quote the identifiers. In Oracle, identifiers are quoted using double quotation marks.
You should choose a schema object name that is unique by case and by at least one other characteristic, and ensure that the object name is not a reserved word from either database.
This section discusses table design issues that you need to consider when converting MySQL databases to Oracle. This section includes the following:
MySQL and Oracle have some differences in the character types that they support and in the way they store and retrieve the character type values.
MySQL supports the CHAR and VARCHAR type for character type with a length that is less than 65,535 bytes. The CHAR type can have a maximum length of 255 bytes, and as of MySQL 3.23 it may also be declared with a length of 0 byte. Before MySQL 5.0.3, the length specification for the VARCHAR type is the same as the CHAR type. From MySQL 5.0.3 on, the maximum length for the VARCHAR type is 65,535 bytes. Oracle supports four character types: CHAR, NCHAR, NVARCHAR2 and VARCHAR2. The minimum length that can be declared for all Oracle character types is 1 byte. The maximum size allowed for CHAR and NCHAR is 2,000 bytes, and for NVARCHAR2 and VARCHAR2 it is 4,000 bytes.
MySQL CHAR values are right-padded with spaces to the specified length when they are stored, and trailing spaces are removed when the values are retrieved. On the other hand, VARCHAR values are stored using as many characters as are given, but before MySQL 5.0.3 trailing spaces are removed when the values are stored and retrieved. Oracle blank-pads the value for its CHAR and NCHAR type to the column length if the value is shorter than the column length, and trailing spaces are not removed on retrieval. For NVARCHAR2 and VARVHAR2 data type columns, Oracle stores and retrieves the value exactly as is given, including trailing spaces.
If a value is assigned to a character type column that exceeds its specified length, MySQL truncates the value and does not generate an error unless the STRICT SQL mode is set. Oracle generates an error if the value assigned to a character type column exceeds its specified length.
In MySQL, every character type (CHAR, VARCHAR, and TEXT) column has a column character set and collation. If the character set or collation is not explicitly defined in the column definition, the table character set or collation is implied if specified; otherwise, the database character or collation is chosen. In Oracle, the character set for CHAR and VARCHAR2 types is defined by the database character set, and for the character set for NCHAR and NVARCHAR types is defined the national character set.
When declaring a CHAR or VARCHAR type in MySQL, the default length semantics is characters instead of bytes for MySQL 4.1 and later. In Oracle, the default length semantics is bytes for CHAR and VARCHAR2 types and characters for NCHAR and NVARCHAR2 types.
SQL Developer will map MySQL CHAR and VARCHAR types to Oracle CHAR and VARCHAR2 types, respectively. SQL Developer will determine the maximum number of bytes for the Oracle CHAR and VARCHAR2 data type columns from the number of bytes required to hold the maximum length specified for the corresponding MySQL CHAR and VARCHAR data type columns. If the MySQL VARCHAR2 column is such that the data exceeds 4000 bytes, convert the column to an Oracle CLOB data type column.
MySQL differs from Oracle in the way it handles default value for a column that does not allow NULL value.
In MySQL, for a column that does not allow NULL value and for which no data is provided for the column when data is inserted into the table, MySQL determines a default value for the column. This default value is the implicit default value for the column data type. However, if the strict mode is enabled, MySQL generates errors, and for transactional tables it rolls back the insert statement.
In Oracle, when data is inserted into a table, data must be provided for all columns that do not allow NULL value. Oracle does not generate a default value for columns that have the NOT NULL constraint.
SQL Developer supports the migration of multiple MySQL databases if they are on the same MySQL database server.
Schema migration considerations for MySQL apply in the following areas:
When migrating MySQL databases to Oracle, SQL Developer maps each MySQL database to a tablespace in Oracle. Database objects, such as tables, indexes and views are stored in the respective tablespaces and are referenced from the Oracle schema for the user that owns them.
SQL Developer does not process all the administrative privileges on MySQL, except the SUPER privilege. Table 2-2 shows the mappings for MySQL per-object privileges granted at the different levels as well as the SUPER privilege granted at the global level.
Table 2-2 MySQL Privileges and Oracle System Privileges
| Level | Privilege | System Privilege(s) on Oracle |
|---|---|---|
|
Global |
ALTER |
ALTER ANY TABLE, ALTER ANY SEQUENCE, ALTER ANY CUSTER, COMMENT ANY TABLE |
|
Global |
ALTER ROUTINE |
ALTER ANY PROCEDURE, DROP ANY PROCEDURE |
|
Global |
CREATE |
CREATE ANY TABLE, CREATE ANY SEQUENCE, CREATE ANY CLUSTER, CREATE DATABASE LINK, COMMENT ANY TABLE |
|
Global |
CREATE ROUTINE |
CREATE ANY PROCEDURE |
|
Global |
CREATE USER |
CREATE USER, GRANT ANY PRIVILEGE |
|
Global |
CREATE VIEW |
CREATE ANY VIEW |
|
Global |
DELETE |
ALTER ANY TABLE, DROP USER, DELETE ANY TABLE |
|
Global |
DROP |
DROP ANY TABLE, DROP ANY SEQUENCE, DROP ANY CLUSTER, DROP ANY VIEW |
|
Global |
EXECUTE |
EXECUTE ANY PROCEDURE |
|
Global |
INDEX |
CREATE ANY INDEX, ALTER ANY INDEX, DROP ANY INDEX |
|
Global |
INSERT |
INSERT ANY TABLE |
|
Global |
LOCK TABLES |
LOCK ANY TABLE |
|
Global |
SELECT |
SELECT ANY TABLE |
|
Global |
SUPER |
CREATE ANY TRIGGER, DROP ANY TRIGGER |
|
Global |
UPDATE |
UPDATE ANY TABLE |
|
Global |
USAGE |
CREATE SESSION, ALTER SESSION, UNLIMITED TABLESPACE |
|
Database |
CREATE |
CREATE CLUSTER, CREATE DATABASE LINK, CREATE SEQUENCE, CREATE TABLE |
|
Database |
CREATE ROUTINE |
CREATE PROCEDURE |
|
Database |
CREATE VIEW |
CREATE VIEW |
|
Table |
CREATE |
CREATE TABLE |
|
Table |
CREATE VIEW |
CREATE VIEW |
SQL Developer does not support the migration of temporary tables.
In MySQL, temporary tables are database objects that are visible only to the current user session and are automatically dropped when the user session ends.
The definition of temporary tables in Oracle differs slightly from MySQL, in that the temporary tables, once created, exist until they are explicitly dropped and they are visible to all sessions with appropriate privileges. However, the data in the temporary tables is visible only to the user session that inserts the data into the table, and the data may persist for the duration of a transaction or a user session.
SQL Developer creates an Oracle schema for the MySQL root user that owns, for all databases to be migrated, all database objects except stored procedures. For stored procedures, the MySQL users that created them remain the owner. SQL Developer creates an Oracle schema for each MySQL user that is migrated.
This section describes the data types used within Oracle. It shows the MySQL data types and the Oracle equivalent. It includes information about the following:
Table 2-3 describes the Oracle data types supported by Oracle SQL Developer.
Table 2-3 Supported Oracle Data Types
| Data Type | Description |
|---|---|
|
BLOB |
A binary large object. Maximum size is 4 gigabytes. |
|
CHAR (SIZE) |
Fixed-length character data of length size bytes. Maximum size is 2000 bytes. Default and minimum size is 1 byte. |
|
CLOB |
A character large object containing single-byte characters. Both fixed-width and variable-width character sets are supported, both using the CHAR database character set. Maximum size is 4 gigabytes. |
|
DATE |
The DATE data type stores date and time information. Although date and time information can be represented in both CHAR and NUMBER data types, the DATE data type has special associated properties. For each DATE value, Oracle stores the following information: century, year, month, day, hour, minute, and second. |
|
FLOAT |
Specifies a floating-point number with decimal precision 38, or binary precision 126. |
|
LONG (SIZE) |
Character data of variable length up to 2 gigabytes, or 2^31 -1 bytes. |
|
LONG RAW |
Raw binary data of variable length up to 2 gigabytes. |
|
NCHAR (SIZE) |
Fixed-length character data of length size characters or bytes, depending on the choice of national character set. Maximum size is determined by the number of bytes required to store each character, with an upper limit of 2000 bytes. Default and minimum size is 1 character or 1 byte, depending on the character set. |
|
NCLOB |
A character large object containing multibyte characters. Both fixed-width and variable-width character sets are supported, both using the NCHAR database character set. Maximum size is 4 gigabytes. Stores national character set data. |
|
NUMBER |
Number having precision p and scale s. The precision p can range from 1 to 38. The scale s can range from -84 to 127. |
|
NVARCHAR2 (SIZE) |
Variable-length character string having maximum length size characters or bytes, depending on the choice of national character set. Maximum size is determined by the number of bytes required to store each character, with an upper limit of 4000 bytes. You must specify size for NVARCHAR2. |
|
RAW (SIZE) |
Raw binary data of length size bytes. Maximum size is 2000 bytes. You must specify size for a RAW value. |
|
VARCHAR (SIZE) |
The VARCHAR data type is currently synonymous with the VARCHAR2 data type. Oracle recommends that you use VARCHAR2 rather than VARCHAR. In the future, VARCHAR might be defined as a separate data type used for variable-length character strings compared with different comparison semantics. The maximum size is 4000 and the minimum of 1 is the default. |
|
BINARY_DOUBLE |
A 64-bit, double-precision floating-point number data type. |
|
BINARY_FLOAT |
A 32-bit, single-precision floating-point number data type. |
For more information about Oracle data types, see Oracle Database SQL Language Reference.
Table 2-4 shows the default settings used by SQL Developer to convert data types from MySQL to Oracle. SQL Developer enables you to change the default setting for certain data types by specifying an alternative type. For information about changing the default data type mappings, see the SQL Developer online help.
Table 2-4 Default Data Type Mappings Used by Oracle SQL Developer
| MySQL Data Type | Oracle Data Type |
|---|---|
|
BIGINT |
NUMBER(19, 0) |
|
BIT |
RAW |
|
BLOB |
BLOB, RAW |
|
CHAR |
CHAR |
|
DATE |
DATE |
|
DATETIME |
DATE |
|
DECIMAL |
FLOAT (24) |
|
DOUBLE |
FLOAT (24) |
|
DOUBLE PRECISION |
FLOAT (24) |
|
ENUM |
VARCHAR2 |
|
FLOAT |
FLOAT |
|
INT |
NUMBER(10, 0) |
|
INTEGER |
NUMBER(10, 0) |
|
LONGBLOB |
BLOB, RAW |
|
LONGTEXT |
CLOB, RAW |
|
MEDIUMBLOB |
BLOB, RAW |
|
MEDIUMINT |
NUMBER(7, 0) |
|
MEDIUMTEXT |
CLOB, RAW |
|
NUMERIC |
NUMBER |
|
REAL |
FLOAT (24) |
|
SET |
VARCHAR2 |
|
SMALLINT |
NUMBER(5, 0) |
|
TEXT |
VARCHAR2, CLOB |
|
TIME |
DATE |
|
TIMESTAMP |
DATE |
|
TINYBLOB |
RAW |
|
TINYINT |
NUMBER(3, 0) |
|
TINYTEXT |
VARCHAR2 |
|
VARCHAR |
VARCHAR2, CLOB |
|
YEAR |
NUMBER |
Note:
The ENUM and SET data types have no direct mapping in Oracle. SQL Developer maps ENUM columns in MySQL to VARCHAR2 columns in Oracle. It then adds a constraint and a trigger to those columns to ensure that only values that were allowed by the ENUM data type are allowed in the column it was mapped to in Oracle.This section lists the difference between MySQL and Oracle data types. For some MySQL data types there is more than one alternative Oracle data type. The tables include information about the following:
When mapping MySQL data types to numeric data types in Oracle, the following conditions apply:
If there is no precision or scale defined for the destination Oracle data type, precision and scale are taken from the MySQL source data type.
If there is a precision or scale defined for the destination data type, these values are compared to the equivalent values of the source data type and the maximum value is selected.
The following table compares the numeric types of MySQL to Oracle:
| MySQL | Size | Oracle |
|---|---|---|
| BIGINT | 8 Bytes | NUMBER (19,0) |
| BIT | approximately (M+7)/8 Bytes | RAW |
| DECIMAL(M,D) | M+2 bytes if D > 0, M+1 bytes if D = 0 (D+2, if M < D) | FLOAT(24), BINARY_FLOAT |
| DOUBLE | 8 Bytes | FLOAT(24), BINARY_FLOAT, BINARY_DOUBLE |
| DOUBLE PRECION | 8 Bytes | FLOAT(24), BINARY_DOUBLE |
| FLOAT(25<=X <=53) | 8 Bytes | FLOAT(24), BINARY_FLOAT |
| FLOAT(X<=24) | 4 Bytes | FLOAT, BINARY_FLOAT |
| INT | 4 Bytes | NUMBER (10,0) |
| INTEGER | 4 Bytes | NUMBER (10,0) |
| MEDIUMINT | 3 Bytes | NUMBER (7,0) |
| NUMERIC | M+2 bytes if D > 0, M+1 bytes if D = 0 (D+2, if M < D) | NUMBER |
| REAL | 8 Bytes | FLOAT(24), BINARY_FLOAT |
| SMALLINT | 2 Bytes | NUMBER(5,0) |
| TINYINT | 1 Byte | NUMBER(3,0) |
The following table compares the date and time types of MySQL to Oracle:
| MySQL | Size | Oracle |
|---|---|---|
| DATE | 3 Bytes | DATE |
| DATETIME | 8 Bytes | DATE |
| TIMESTAMP | 4 Bytes | DATE |
| TIME | 3 Bytes | DATE |
| YEAR | 1 Byte | NUMBER |
When mapping MySQL data types to character data types in Oracle, the following conditions apply:
If there is no length defined for the destination data type, the length is taken from the source data type.
If there is a length defined for the destination data type, the maximum value of the two lengths is taken.
The following table compares the string types of MySQL to Oracle:
Note:
Reference to M indicates the maximum display size. The maximum legal display size is 255. A reference to L applies to a floating point types and indicates the number of digits following the decimal point.| MySQL | Size | Oracle |
|---|---|---|
| BLOB | L + 2 Bytes whereas L<2^16 | RAW, BLOB |
| CHAR(m) | M Bytes, 0<=M<=255 | CHAR |
| ENUM (VALUE1, VALUE2, ...) | 1 or 2 Bytes depending on the number of enum. values (65535 values max) | |
| LONGBLOB | L + 4 Bytes whereas L < 2 ^ 32 | RAW, BLOB |
| LONGTEXT | L + 4 Bytes whereas L < 2 ^ 32 | RAW, CLOB |
| MEDIUMBLOB | L + 3 Bytes whereas L < 2^ 24 | RAW, BLOB |
| MEDIUMTEXT | L + 3 Bytes whereas L < 2^ 24 | RAW, CLOB |
| SET (VALUE1, VALUE2, ...) | 1, 2, 3, 4 or 8 Bytes depending on the number of set members (64 members maximum) | |
| TEXT | L + 2 Bytes whereas L<2^16 | VARCHAR2, CLOB |
| TINYBLOB | L + 1 Bytes whereas L<2 ^8 | RAW, BLOB |
| TINYTEXT | L + 1 Bytes whereas L<2 ^8 | VARCHAR2 |
| VARCHAR(m) | L+1 Bytes whereas L<=M and0<=M<=255 before MySQL 5.0.3 (0 <= M <= 65535 in MySQL 5.0.3 and later; effective maximum length is 65,532 bytes) | VARCHAR2, CLOB |
This section provide a description of the conceptual differences and similarities in data storage for MySQL and Oracle databases.
Data storage is an aspect of MySQL that sets it apart for nearly every database, including Oracle. In MySQL, databases correspond to directories within the data directory of the server. Tables within a database correspond to one or more files within the database directory, depending on the storage engine used for the tables.
A database can contain a mix of tables of different storage engines. A storage engine is responsible for the storage and retrieval of the data for a table.
MySQL offers a variety of storage engines (formerly called table types) to meet the different requirements of the user's environment. Table 2-5 shows the storage engines supported by MySQL.
Table 2-5 Storage Engines Supported by MySQL
| Storage Engine | Description |
|---|---|
|
MyISAM |
The default non-transactional storage engine that provides full-text indexing and is highly portable |
|
MERGE |
A non-transactional storage engine that allows a collection of MyISAM tables with identical column and index information to be used as one |
|
MEMORY (HEAP) |
A non-transactional storage engine that stores data in memory |
|
BDB (Berkeley DB) |
The first transactional-safe storage engine |
|
InnoDB |
A transactional-safe storage engine designed for maximum performance when processing large volume of data and that provides row-level locking |
|
FEDERATED |
A storage engine that accesses data in tables of remote databases rather than in local tables |
|
ARCHIVE |
A storage engine that can store large amount of data without indexes in very small footprint |
|
CSV |
A storage engine that stores data in text file using comma-separated-values format |
|
BLACKHOLE |
A storage engine that acts as a "black hole" that accepts data but throws it away and does not store it |
|
EXAMPLE |
A "stub" engine that does nothing. Its purpose is to serve as an example that illustrates how to begin writing new engines. |
|
ISAM |
The original MySQL storage engine that has been deprecated in favor of the MyISAM storage engine as of version 5.0 |
Each storage engine has its benefits and drawbacks. Some of features that differentiate the storage engines are transaction, locking, concurrency and portability. The following table summarizes the features for four of the commonly used storage engines.
Table 2-6 Feature Comparison for Common Storage Engines
| Feature | MyISAM | Heap | BDB | InnoDB |
|---|---|---|---|---|
|
Transactional |
No |
No |
Yes |
Yes |
|
Lock granularity |
Table |
Table |
Page |
Row |
|
Storage |
A data file (.MYD) and an index file (.MYI) for each table |
In-memory |
A single data and index file (.db) for each table |
A set of data files for all the tables |
|
Portable |
Yes |
N/A |
No |
Yes |
An Oracle database consists of one or more tablespaces. Tablespaces provide logical storage space that link a database to the physical disks that hold the data. A tablespace is created from one or more data files. Data files are files in the file system or an area of disk space specified by a raw device. A tablespace can be enlarged by adding more data files.
An Oracle database consists of a least a SYSTEM tablespace, where the Oracle tables are stored. It can also consist of user defined tablespaces. A tablespace is the logical storage location for database objects. For example, you can specify where a particular table or index gets created in the tablespace.