MySQL now provides a method for storing authentication credentials encrypted in an option file named .mylogin.cnf. To create the file, use the mysql_config_editor utility. The file can be read later by MySQL client programs to obtain authentication credentials for connecting to a MySQL server. mysql_config_editor writes the .mylogin.cnf file using encryption so the credentials are not stored as clear text, and its contents when decrypted by client programs are used only in memory. In this way, passwords can be stored in a file in non-cleartext format and used later without ever needing to be exposed on the command line or in an environment variable. For more information, see Section 4.6.6, “mysql_config_editor — MySQL Configuration Utility”.

MySQL now supports stronger encryption for user account passwords, available through an authentication plugin named sha256_password that implements SHA-256 password hashing. This plugin is built in, so it is always available and need not be loaded explicitly. For more information, including instructions for creating accounts that use SHA-256 passwords, see Section 6.3.8.4, “The SHA-256 Authentication Plugin”.

The mysql.user table now has a password_expired column. Its default value is 'N', but can be set to 'Y' with the new ALTER USER statement. After an account's password has been expired, all operations performed in subsequent connections to the server using the account result in an error until the user issues a SET PASSWORD statement to establish a new account password. For more information, see Section 13.7.1.1, “ALTER USER Syntax”, and Section 6.3.6, “Password Expiration and Sandbox Mode”.

MySQL now has provision for checking password security:

Both capabilities are implemented by the validate_password plugin. For more information, see Section 6.1.2.6, “The Password Validation Plugin”.

mysql_upgrade now produces a warning if it finds user accounts with passwords hashed with the older pre-4.1 hashing method. Such accounts should be updated to use more secure password hashing. See Section 6.1.2.4, “Password Hashing in MySQL”

On Unix platforms, mysql_install_db supports a new option, --random-passwords, that provides for more secure MySQL installation. Invoking mysql_install_db with --random-passwords causes it to assign a random password to the MySQL root accounts, set the “password expired” flag for those accounts, and remove the anonymous-user MySQL accounts. For additional details, see Section 4.4.3, “mysql_install_db — Initialize MySQL Data Directory”.

Logging has been modified so that passwords do not appear in plain text in statements written to the general query log, slow query log, and binary log. See Section 6.1.2.3, “Passwords and Logging”.

The mysql client no longer logs to its history file statements that refer to passwords. See Section 4.5.1.3, “mysql Logging”.

START SLAVE syntax has been modified to permit connection parameters to be specified for connecting to the master. This provides an alternative to storing the password in the master.info file. See Section 13.4.2.5, “START SLAVE Syntax”.

You can create FULLTEXT indexes on InnoDB tables, and query them using the MATCH() ... AGAINST syntax. This feature includes a new proximity search operator (@) and several new configuration options and INFORMATION_SCHEMA tables: See Section 14.2.2.13.3, “FULLTEXT Indexes” for more information.

Several ALTER TABLE operations can be performed without copying the table, without blocking inserts, updates, and deletes to the table, or both. These enhancements are known collectively as online DDL. See Section 14.2.11, “InnoDB and Online DDL” for details.

InnoDB now supports the DATA DIRECTORY='directory' clause of the CREATE TABLE statement, which allows you to create InnoDB file-per-table tablespaces (.ibd files) in a location outside the MySQL data directory. This enhancement provides the flexibility to create file-per-table tablespaces in locations that better suit your server environment. For example, you could place busy tables on an SSD device, or large tables on a high-capacity HDD device.

For additional information, see Section 14.2.5.4, “Specifying the Location of a Tablespace”.

InnoDB now supports the notion of “transportable tablespaces”, allowing file-per-table tablespaces (.ibd files) to be exported from a running MySQL instance and imported into another running instance without inconsistencies or mismatches caused by buffered data, in-progress transactions, and internal bookkeeping details such as the space ID and LSN.

The FOR EXPORT clause of the FLUSH TABLE command writes any unsaved changes from InnoDB memory buffers to the .ibd file. After copying the .ibd file and a separate metadata file to the other server, the DISCARD TABLESPACE and IMPORT TABLESPACE clauses of the ALTER TABLE statement are used to bring the table data into a different MySQL instance.

This enhancement provides the flexibility to move file-per-table tablespaces around to better suit your server environment. For example, you could move busy tables to an SSD device, or move large tables to a high-capacity HDD device. For more information, see Section 14.2.5.5, “Copying Tablespaces to Another Server (Transportable Tablespaces)”.

You can now set the InnoDB page size for uncompressed tables to 8KB or 4KB, as an alternative to the default 16KB. This setting is controlled by the innodb_page_size configuration option. You specify the size when creating the MySQL instance. All InnoDB tablespaces within an instance share the same page size. Smaller page sizes can help to avoid redundant or inefficient I/O for certain combinations of workload and storage devices, particularly SSD devices with small block sizes.

Improvements to the algorithms for adaptive flushing make I/O operations more efficient and consistent under a variety of workloads. The new algorithm and default configuration values are expected to improve performance and concurrency for most users. Advanced users can fine-tune their I/O responsiveness through several configuration options. See Section 14.2.12.2.9, “Improvements to Buffer Pool Flushing” for details.

You can code MySQL applications that access InnoDB tables through a NoSQL-style API. This feature uses the popular memcached daemon to relay requests such as ADD, SET, and GET for key-value pairs. These simple operations to store and retrieve data avoid the SQL overhead such as parsing and constructing a query execution plan. You can access the same data through the NoSQL API and SQL. For example, you might use the NoSQL API for fast updates and lookups, and SQL for complex queries and compatibility with existing applications. See Section 14.2.16, “InnoDB Integration with memcached” for details.

Optimizer statistics for InnoDB tables are gathered at more predictable intervals and can persist across server restarts, for improved plan stability. You can also control the amount of sampling done for InnoDB indexes, to make the optimizer statistics more accurate and improve the query execution plan. See Section 14.2.12.2.10, “Persistent Optimizer Statistics for InnoDB Tables” for details.

New optimizations apply to read-only transactions, improving performance and concurrency for ad-hoc queries and report-generating applications. These optimizations are applied automatically when practical, or you can specify START TRANSACTION READ ONLY to ensure the transaction is read-only. See Section 14.2.12.2.3, “Optimizations for Read-Only Transactions” for details.

You can move the InnoDB undo log out of the system tablespace into one or more separate tablespaces. The I/O patterns for the undo log make these new tablespaces good candidates to move to SSD storage, while keeping the system tablespace on hard disk storage. For details, see Section 14.2.12.2.4, “Separate Tablespaces for InnoDB Undo Logs”.

You can improve the efficiency of the InnoDB checksum feature by specifying the configuration option innodb_checksum_algorithm=crc32, which turns on a faster checksum algorithm. This option replaces the innodb_checksums option. Data written using the old checksum algorithm (option value innodb) is fully upward-compatible; tablespaces modified using the new checksum algorithm (option value crc32) cannot be downgraded to an earlier version of MySQL that does not support the innodb_checksum_algorithm option.

The InnoDB redo log files now have a maximum combined size of 512GB, increased from 4GB. You can specify the larger values through the innodb_log_file_size option. The startup behavior now automatically handles the situation where the size of the existing redo log files does not match the size specified by innodb_log_file_size and innodb_log_files_in_group.

The --innodb-read-only option lets you run a MySQL server in read-only mode. You can access InnoDB tables on read-only media such as a DVD or CD, or set up a data warehouse with multiple instances all sharing the same data directory. See Section 14.2.3.1, “Configuring InnoDB for Read-Only Operation” for usage details.

A new configuration option, innodb_compression_level, allows you to select a compression level for InnoDB compressed tables, from the familiar range of 0-9 used by zlib. You can also control whether compressed pages in the buffer pool are stored in the redo log when an update operation causes pages to be compressed again. This behavior is controlled by the innodb_log_compressed_pages configuration option.

Data blocks in an InnoDB compressed table contain a certain amount of empty space (padding) to allow DML operations to modify the row data without re-compressing the new values. Too much padding can increase the chance of a compression failure, requiring a page split, when the data does need to be re-compressed after extensive changes. The amount of padding can now be adjusted dynamically, so that DBAs can reduce the rate of compression failures without re-creating the entire table with new parameters, or re-creating the entire instance with a different page size. The associated new configuration options are innodb_compression_failure_threshold_pct, innodb_compression_pad_pct_max.

Several new InnoDB-related INFORMATION_SCHEMA tables provide information about the InnoDB buffer pool, metadata about tables, indexes, and foreign keys from the InnoDB data dictionary, and low-level information about performance metrics that complements the information from the Performance Schema tables.

To ease the memory load on systems with huge numbers of tables, InnoDB now frees up the memory associated with an opened table using an LRU algorithm to select tables that have gone the longest without being accessed. To reserve more memory to hold metadata for open InnoDB tables, increase the value of the table_definition_cache configuration option. InnoDB treats this value as a “soft limit” for the number of open table instances in the InnoDB data dictionary cache. For additional information, refer to the table_definition_cache documentation.

InnoDB has several internal performance enhancements, including reducing contention by splitting the kernel mutex, moving flushing operations from the main thread to a separate thread, enabling multiple purge threads, and reducing contention for the buffer pool on large-memory systems.

InnoDB uses a new, faster algorithm to detect deadlocks. Information about all InnoDB deadlocks can be written to the MySQL server error log, to help diagnose application issues.

To avoid a lengthy warmup period after restarting the server, particularly for instances with large InnoDB buffer pools, you can reload pages into the buffer pool immediately after a restart. MySQL can dump a compact data file at shutdown, then consult that data file to find the pages to reload on the next restart. You can also manually dump or reload the buffer pool at any time, for example during benchmarking or after complex report-generation queries. See Section 14.2.12.2.8, “Faster Restart by Preloading the InnoDB Buffer Pool” for details.

As of MySQL 5.6.16, innochecksum provides support for files greater than 2GB in size. Previously, innochecksum only supported files up to 2GB in size.

As of MySQL 5.6.16, new global configuration parameters, innodb_status_output and innodb_status_output_locks, allow you to dynamically enable and disable the standard InnoDB Monitor and InnoDB Lock Monitor for periodic output. Enabling and disabling monitors for periodic output by creating and dropping specially named tables is deprecated and may be removed in a future release. For additional information, see Section 14.2.12.4, “InnoDB Monitors”.

As of MySQL 5.6.17, MySQL supports rebuilding regular and partitioned InnoDB tables using online DDL (ALGORITHM=INPLACE) for the following operations:

Online DDL support reduces table rebuild time and permits concurrent DML, which helps reduce user application downtime. For additional information, see Section 14.2.11.1, “Overview of Online DDL”.

The maximum number of partitions is increased to 8192. This number includes all partitions and all subpartitions of the table.

It is now possible to exchange a partition of a partitioned table or a subpartition of a subpartitioned table with a nonpartitioned table that otherwise has the same structure using the ALTER TABLE ... EXCHANGE PARTITION statement. This can be used, for example, to import and export partitions. For more information and examples, see Section 18.3.3, “Exchanging Partitions and Subpartitions with Tables”.

Explicit selection of one or more partitions or subpartitions is now supported for queries, as well as for many data modification statements, that act on partitioned tables. For example, assume a table t with some integer column c has 4 partitions named p0, p1, p2, and p3. Then the query SELECT * FROM t PARTITION (p0, p1) WHERE c < 5 returns only those rows from partitions p0 and p1 for which c is less than 5.

The following statements support explicit partition selection:

For syntax, see the descriptions of the individual statements. For additional information and examples, see Section 18.5, “Partition Selection”.

Partition lock pruning greatly improves performance of many DML and DDL statements acting on tables with many partitions by helping to eliminate locks on partitions that are not affected by these statements. Such statements include many SELECT, SELECT ... PARTITION, UPDATE, REPLACE, INSERT, as well as many other statements. For more information, including a complete listing of the statements whose performance has thus been improved, see Section 18.6.4, “Partitioning and Locking”.

Instrumentation for table input and output. Instrumented operations include row-level accesses to persistent base tables or temporary tables. Operations that affect rows are fetch, insert, update, and delete.

Event filtering by table, based on schema and/or table names.

Event filtering by thread. More information is collected for threads.

Summary tables for table and index I/O, and for table locks.

Instrumentation for statements and stages within statements.

Configuration of instruments and consumers at server startup, which previously was possible only at runtime.

MySQL now supports transaction-based replication using global transaction identifiers (also known as “GTIDs”). This makes it possible to identify and track each transaction when it is committed on the originating server and as it is applied by any slaves.

Enabling of GTIDs in a replication setup is done primarily using the new --gtid-mode and --enforce-gtid-consistency server options. For information about additional options and variables introduced in support of GTIDs, see Section 16.1.4.5, “Global Transaction ID Options and Variables”.

When using GTIDs it is not necessary to refer to log files or positions within those files when starting a new slave or failing over to a new master, which greatly simplifies these tasks. For more information about provisioning servers for GTID replication with or without referring to binary log files, see Section 16.1.3.3, “Using GTIDs for Failover and Scaleout”.

GTID-based replication is completely transaction-based, which makes it simple to check the consistency of masters and slaves. If all transactions committed on a given master are also committed on a given slave, consistency between the two servers is guaranteed.

For more complete information about the implementation and use of GTIDs in MySQL Replication, see Section 16.1.3, “Replication with Global Transaction Identifiers”.

MySQL row-based replication now supports row image control. By logging only those columns required for uniquely identifying and executing changes on each row (as opposed to all columns) for each row change, it is possible to save disk space, network resources, and memory usage. You can determine whether full or minimal rows are logged by setting the binlog_row_image server system variable to one of the values minimal (log required columns only), full (log all columns), or noblob (log all columns except for unneeded BLOB or TEXT columns). See System variables used with the binary log, for more information.

Binary logs written and read by the MySQL Server are now crash-safe, because only complete events (or transactions) are logged or read back. By default, the server logs the length of the event as well as the event itself and uses this information to verify that the event was written correctly. You can also cause the server to write checksums for the events using CRC32 checksums by setting the binlog_checksum system variable. To cause the server to read checksums from the binary log, use the master_verify_checksum system variable. The --slave-sql-verify-checksum system variable causes the slave SQL thread to read checksums from the relay log.

MySQL now supports logging of master connection information and of slave relay log information to tables as well as files. Use of these tables can be controlled independently, by the --master-info-repository and --relay-log-info-repository server options. Setting --master-info-repository to TABLE causes connection information to be logged in the slave_master_info table; setting --relay-log-info-repository to TABLE causes relay log information to be logged to the slave_relay_log_info table. Both of these tables are created automatically, in the mysql system database.

In order for replication to be crash-safe, the slave_master_info and slave_relay_log_info tables must each use a transactional storage engine, and beginning with MySQL 5.6.6, these tables are created using InnoDB for this reason. (Bug #13538891) If you are using a previous MySQL 5.6 release in which both of these tables use MyISAM, this means that, prior to starting replication, you must convert both of them to a transactional storage engine (such as InnoDB) if you wish for replication to be crash-safe. You can do this in such cases by means of the appropriate ALTER TABLE ... ENGINE=... statements. You should not attempt to change the storage engine used by either of these tables while replication is actually running.

mysqlbinlog now has the capability to back up a binary log in its original binary format. When invoked with the --read-from-remote-server and --raw options, mysqlbinlog connects to a server, requests the log files, and writes output files in the same format as the originals. See Section 4.6.8.3, “Using mysqlbinlog to Back Up Binary Log Files”.

MySQL now supports delayed replication such that a slave server deliberately lags behind the master by at least a specified amount of time. The default delay is 0 seconds. Use the new MASTER_DELAY option for CHANGE MASTER TO to set the delay.

Delayed replication can be used for purposes such as protecting against user mistakes on the master (a DBA can roll back a delayed slave to the time just before the disaster) or testing how the system behaves when there is a lag. See Section 16.3.9, “Delayed Replication”.

A replication slave having multiple network interfaces can now be caused to use only one of these (to the exclusion of the others) by using the MASTER_BIND option when issuing a CHANGE MASTER TO statement.

The log_bin_basename system variable has been added. This variable contains the complete filename and path to the binary log file. Whereas the log_bin system variable shows only whether or not binary logging is enabled, log_bin_basename reflects the name set with the --log-bin server option.

Similarly, the relay_log_basename system variable shows the filename and complete path to the relay log file.

MySQL Replication now supports parallel execution of transactions with multi-threading on the slave. When parallel execution is enabled, the slave SQL thread acts as the coordinator for a number of slave worker threads as determined by the value of the slave_parallel_workers server system variable. The current implementation of multi-threading on the slave assumes that data and updates are partitioned on a per-database basis, and that updates within a given database occur in the same relative order as they do on the master. However, it is not necessary to coordinate transactions between different databases. Transactions can then also be distributed per database, which means that a worker thread on the slave slave can process successive transactions on a given database without waiting for updates to other databases to complete.

Since transactions on different databases can occur in a different order on the slave than on the master, simply checking for the most recently executed transaction is not a guarantee that all previous transactions on the master have been executed on the slave. This has implications for logging and recovery when using a multi-threaded slave. For information about how to interpret binary logging information when using multi-threading on the slave, see Section 13.7.5.35, “SHOW SLAVE STATUS Syntax”.

The optimizer now more efficiently handles queries (and subqueries) of the following form:

SELECT ... FROM single_table ... ORDER BY non_index_column [DESC] LIMIT [M,]N;

That type of query is common in web applications that display only a few rows from a larger result set. For example:

SELECT col1, ... FROM t1 ... ORDER BY name LIMIT 10;
SELECT col1, ... FROM t1 ... ORDER BY RAND() LIMIT 15;

The sort buffer has a size of sort_buffer_size. If the sort elements for N rows are small enough to fit in the sort buffer (M+N rows if M was specified), the server can avoid using a merge file and perform the sort entirely in memory. For details, see Section 8.2.1.19, “Optimizing LIMIT Queries”.

The optimizer implements Disk-Sweep Multi-Range Read. Reading rows using a range scan on a secondary index can result in many random disk accesses to the base table when the table is large and not stored in the storage engine's cache. With the Disk-Sweep Multi-Range Read (MRR) optimization, MySQL tries to reduce the number of random disk access for range scans by first scanning the index only and collecting the keys for the relevant rows. Then the keys are sorted and finally the rows are retrieved from the base table using the order of the primary key. The motivation for Disk-sweep MRR is to reduce the number of random disk accesses and instead achieve a more sequential scan of the base table data. For more information, see Section 8.2.1.13, “Multi-Range Read Optimization”.

The optimizer implements Index Condition Pushdown (ICP), an optimization for the case where MySQL retrieves rows from a table using an index. Without ICP, the storage engine traverses the index to locate rows in the base table and returns them to the MySQL server which evaluates the WHERE condition for the rows. With ICP enabled, and if parts of the WHERE condition can be evaluated by using only fields from the index, the MySQL server pushes this part of the WHERE condition down to the storage engine. The storage engine then evaluates the pushed index condition by using the index entry and only if this is satisfied is base row be read. ICP can reduce the number of accesses the storage engine has to do against the base table and the number of accesses the MySQL server has to do against the storage engine. For more information, see Section 8.2.1.6, “Index Condition Pushdown Optimization”.

The EXPLAIN statement now provides execution plan information for DELETE, INSERT, REPLACE, and UPDATE statements. Previously, EXPLAIN provided information only for SELECT statements. In addition, the EXPLAIN statement now can produce output in JSON format. See Section 13.8.2, “EXPLAIN Syntax”.

The optimizer more efficiently handles subqueries in the FROM clause (that is, derived tables). Materialization of subqueries in the FROM clause is postponed until their contents are needed during query execution, which improves performance. In addition, during query execution, the optimizer may add an index to a derived table to speed up row retrieval from it. For more information, see Section 8.2.1.18.3, “Optimizing Subqueries in the FROM Clause (Derived Tables)”.

The optimizer uses semi-join and materialization strategies to optimize subquery execution. See Section 8.2.1.18.1, “Optimizing Subqueries with Semi-Join Transformations”, and Section 8.2.1.18.2, “Optimizing Subqueries with Subquery Materialization”.

A Batched Key Access (BKA) join algorithm is now available that uses both index access to the joined table and a join buffer. The BKA algorithm supports inner join, outer join, and semi-join operations, including nested outer joins and nested semi-joins. Benefits of BKA include improved join performance due to more efficient table scanning. For more information, see Section 8.2.1.14, “Block Nested-Loop and Batched Key Access Joins”.

The optimizer now has a tracing capability, primarily for use by developers. The interface is provided by a set of optimizer_trace_xxx system variables and the INFORMATION_SCHEMA.OPTIMIZER_TRACE table. For details, see MySQL Internals: Tracing the Optimizer.

Block scope is used in determining which handler to select. Previously, a stored program was treated as having a single scope for handler selection.

Condition precedence is more accurately resolved.

Diagnostics area clearing has changed. Bug #55843 caused handled conditions to be cleared from the diagnostics area before activating the handler. This made condition information unavailable within the handler. Now condition information is available to the handler, which can inspect it with the GET DIAGNOSTICS statement. The condition information is cleared when the handler exits, if it has not already been cleared during handler execution.

Previously, handlers were activated as soon as a condition occurred. Now they are not activated until the statement in which the condition occurred finishes execution, at which point the most appropriate handler is chosen. This can make a difference for statements that raise multiple conditions, if a condition raised later during statement execution has higher precedence than an earlier condition and there are handlers in the same scope for both conditions. Previously, the handler for the first condition raised would be chosen, even if it had a lower precedence than other handlers. Now the handler for the condition with highest precedence is chosen, even if it is not the first condition raised by the statement.

MySQL now permits fractional seconds for TIME, DATETIME, and TIMESTAMP values, with up to microseconds (6 digits) precision. See Section 11.3.6, “Fractional Seconds in Time Values”.

Previously, at most one TIMESTAMP column per table could be automatically initialized or updated to the current date and time. This restriction has been lifted. Any TIMESTAMP column definition can have any combination of DEFAULT CURRENT_TIMESTAMP and ON UPDATE CURRENT_TIMESTAMP clauses. In addition, these clauses now can be used with DATETIME column definitions. For more information, see Section 11.3.5, “Automatic Initialization and Updating for TIMESTAMP and DATETIME”.

In MySQL, the TIMESTAMP data type differs in nonstandard ways from other data types in terms of default value and assignment of automatic initialization and update attributes. These behaviors remain the default but now are deprecated, and can be turned off by enabling the explicit_defaults_for_timestamp system variable at server startup. See Section 11.3.5, “Automatic Initialization and Updating for TIMESTAMP and DATETIME”, and Section 5.1.4, “Server System Variables”.

New Connection_errors_xxx status variables provide information about connection errors that do not apply to specific client IP addresses.

Counters have been added to the host cache to track errors that do apply to specific IP addresses, and a new host_cache Performance Schema table exposes the contents of the host cache so that it can be examined using SELECT statements. Access to host cache contents makes it possible to answer questions such as how many hosts are cached, what kinds of connection errors are occurring for which hosts, or how close host error counts are to reaching the max_connect_errors system variable limit.

The host cache size now is configurable using the host_cache_size system variable.