MySQL 5.5

MySQL 5.5 and Scalability
Ligaya Turmelle
Senior Technical Support Engineer - MySQL
[email_address]
http://joind.in/1584
<Insert Picture Here>

The following is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle’s products remains at the sole discretion of Oracle.

Areas to Cover
Semi-Synchronous Replication
Performance Schema
SIGNAL/RESIGNAL
More Partitioning Options
InnoDB - LOTS of InnoDB!
Performance Improvements
Scalability Improvements
<Insert Picture Here>

Semi-Synchronous Replication http://dev.mysql.com/doc/refman/5.5/en/replication-semisync.html <Insert Picture Here>

History
By default is asynchronous
Master writes to binary log
Slave connects and “pulls” contents of the binary log
Master crashes
No guarantee that a slave has all committed transactions
Potential lost transactions
Failover to slave
May be missing transactions from master
Loss of data integrity

What is Semi-Synchronous Replication?
Alternative to asynchronous replication
Midway point between asynchronous and fully synchronous
Doesn’t have to wait for all the slaves to *receive* the event
The master blocks, waiting for a single slave to acknowledge the commit or timeout
Doesn’t have to wait for all slaves to *commit* the transaction
The slave acknowledges after it receives all the events for the transaction, write it to its relay logs, and flush to disk
Provides improved data integrity at the cost of performance

How does it works
Best Case
COMMIT
Master blocks waiting for acknowledgment
On acknowledgement the master returns to the session
Worst Case
COMMIT
Master blocks
Timeout
Master reverts to asynchronous mode until at least one semi-synchronous slave catches up
Revert back to semi-sychronous mode

Performance Schema http://dev.mysql.com/doc/refman/5.5/en/performance-schema.html <Insert Picture Here>

Basic Info
Monitoring MySQL Server execution at a *very* low level.
Used to inspect internal execution of the server at run time
Focuses on performance data
Uses the Performance Schema storage engine and the Performance_Schema database
Tables in Performance_Schema are views or temporary tables.
Available on all platforms but some limitations might apply
Activation causes no change in server behavior
Average user has no need to mess with it - designed for advanced users

SIGNAL/RESIGNAL http://dev.mysql.com/doc/refman/5.5/en/signal-resignal.html <Insert Picture Here>

SIGNAL
Way to “return” an error.
AKA exception-handling logic for stored procedures, stored functions, triggers, events, and database applications - think throw
Allows you to control error number, SQLSTATE value, message
Can indicate errors, warnings, or “not found.”
Requires no special permissions

CREATE PROCEDURE p (pval INT) BEGIN DECLARE specialty CONDITION FOR SQLSTATE '45000'; IF pval = 0 THEN SIGNAL SQLSTATE '01000'; ELSEIF pval = 1 THEN SIGNAL SQLSTATE '45000' SET MESSAGE_TEXT = 'An error occurred'; ELSEIF pval = 2 THEN SIGNAL specialty SET MESSAGE_TEXT = 'An error occurred'; ELSE SIGNAL SQLSTATE '01000' SET MESSAGE_TEXT = 'A warning occurred', MYSQL_ERRNO = 1000; SIGNAL SQLSTATE '45000' SET MESSAGE_TEXT = 'An error occurred', MYSQL_ERRNO = 1001; END IF; END;

RESIGNAL
passes on error condition information available during execution of a handler.
Requires an active handler to execute
Possible to both handle an error and return the error information. Think catch/throw block.
May change some or all information before passing it on.
No special permissions needed to use
Multiple possible uses
Alone
With new signal information
With a condition value and possibly new signal information
Otherwise follows the rules of SIGNAL

RESIGNAL alone
Pass on error with no change
Allows for the performance of additional actions in the active handler, and then passing on the error message without changing the original condition information
CREATE PROCEDURE p () BEGIN DECLARE EXIT HANDLER FOR SQLEXCEPTION BEGIN SET @error_count = @error_count + 1; IF @a = 0 THEN RESIGNAL; END IF; END; DROP TABLE xx; END

RESIGNAL with New Signal Information
Pass on error with changes
Allows for the performance of additional actions in the active handler, and can change any or all of the signal information items
CREATE PROCEDURE p () BEGIN DECLARE EXIT HANDLER FOR SQLEXCEPTION BEGIN SET @error_count = @error_count + 1; IF @a = 0 THEN RESIGNAL SET MYSQL_ERRNO = 5; END IF; END; DROP TABLE xx; END

RESIGNAL with a Condition Value and Possibly New Signal Information
Push a new “condition” into the current error stack
Similar effects of previous example, except instead of rewriting the signal information it adds a new signal to the top of the error stack
CREATE PROCEDURE p () BEGIN DECLARE EXIT HANDLER FOR SQLEXCEPTION BEGIN SET @error_count = @error_count + 1; IF @a = 0 THEN RESIGNAL SQLSTATE '45000' SET MYSQL_ERRNO=5; END IF; END; DROP TABLE xx; END

Column Partitioning Options http://dev.mysql.com/doc/refman/5.5/en/partitioning-columns.html <Insert Picture Here>

Column Partitioning
Variants on RANGE and LIST partitioning
RANGE COLUMNS
LIST COLUMNS
Allows the use of multiple columns in partitioning keys
All columns taken into account
for placing rows in partitions
for use in partition pruning.
Support the use of non-integer columns
All integer types
DATE and DATETIME
CHAR , VARCHAR , BINARY , and VARBINARY
Great article at http://dev.mysql.com/tech-resources/articles/mysql_55_partitioning.html

Range Column
Allows for multiple column values
Major differences from just RANGE
Does not accept expressions, only names of columns
Accepts a list of one or more columns
Not restricted to integer columns
Go to the example

CREATE TABLE rc1 (a INT, b INT) PARTITION BY RANGE COLUMNS(a, b) ( PARTITION p0 VALUES LESS THAN (5, 12), PARTITION p3 VALUES LESS THAN (MAXVALUE, MAXVALUE) ); mysql> INSERT INTO rc1 VALUES (5,10), (5,11), (5,12); Query OK, 3 rows affected (0.00 sec) Records: 3 Duplicates: 0 Warnings: 0 mysql> SELECT PARTITION_NAME,TABLE_ROWS -> FROM INFORMATION_SCHEMA.PARTITIONS -> WHERE TABLE_NAME = 'rc1'; +--------------+----------------+------------+ | TABLE_SCHEMA | PARTITION_NAME | TABLE_ROWS | +--------------+----------------+------------+ | p | p0 | 2 | | p | p1 | 1 | +--------------+----------------+------------+

List Column
Again allows for multiple column values
Allows you to use
string types
DATE ,
DATETIME
Do not need to convert values to integers to work with
Much easier to read
Go to the example

CREATE TABLE customers_1 ( first_name VARCHAR(25), last_name VARCHAR(25), street_1 VARCHAR(30), city VARCHAR(15), renewal DATE ) PARTITION BY LIST COLUMNS(city) ( PARTITION NorthEast VALUES IN('Boston', 'New York', 'Providence'), PARTITION South VALUES IN('Miami', 'Atlanta', 'New Orleans'), PARTITION MidWest VALUES IN('Chicago', 'Houston', 'Denver'), PARTITION West VALUES IN('Los Angeles', 'San Francisco', 'Seattle') );

InnoDB http://dev.mysql.com/doc/innodb-plugin/1.1/en/innodb-performance.html <Insert Picture Here>

System Mutex Library
Originally used its own implementation of mutexes and read/write locks with Pthreads
From 1.0.3, changed to the GNU Compiler Collection (GCC) for atomic memory access instead of using the Pthreads approach previously used.
Faster more efficient locking
Greater scalability on multi-core platforms.
Enabled out of the box on platforms where supported
Freebie

Memory Allocator
History:
Originally built its own and protected by a single mutex
Also included a wrapper around the system allocator (malloc and free)
also protected with a single mutex
All bad for multi-core computers
Great improvements have occurred since then
Perform better and are more scalable than they were in the past
Most workloads benefit from using the new schedulers.
innodb_use_sys_malloc

Change Buffer
Most relevant to IO bound systems
Will cache changes to secondary index entries
Can buffer inserts, delete-marking operations, and purges
But...
Resides in buffer pool - reducing memory to cache data
Increases risk of a purge hold
Can control whether InnoDB performs this buffering using innodb_change_buffering.
Multiple values available to buffer only what you want

Read Ahead
Prefetch multiple pages in the buffer cache asynchronously
Only the Linear algorithm now
Based on access pattern of pages, not just their number
Can now control when InnoDB performs a read-ahead operation with innodb_read_ahead_threshold
Higher the value, the more strict the access pattern
Efficiency can be found with Innodb_buffer_pool_read_ahead_evicted < Innodb_buffer_pool_read_ahead

Background IO Threads
History
1 read, 1 write, 1 log, 1 insert buffer (non-Windows)
Windows used innodb_file_io_threads (deprecated)
Now configurable
Number of background *helper* threads tasked with servicing any read and write I/O on the data pages
Makes InnoDB more scalable on high end systems

INNODB_IO_CAPACITY
Controls the IO capacity of the InnoDB Master thread.
It controls how many dirty pages are flushed and the size of the ibuf contraction
Also handles the purge but...
Should be set to approximately the number of I/O operations that the system can perform per second

Improved Purge Scheduling
History
One of the operations handled by the Master Thread
Now moved to its own thread
separates main database operations from maintenance work happening in the background.
Improves scalability
May not see improvement on speed since it may hit new types of contention
lays groundwork for further tuning and potentially more then one purge thread.
Enable with innodb_purge_threads=1

Adaptive Dirty Page Flushing
History
Performed in background
Performed by master thread
Currently will aggressively flush buffer pool pages if innodb_max_dirty_pages_pct percentage exceeded
New algorithm
Intent is to smooth overall performance, eliminating steep dips in throughput
Function of the number of dirty pages in the buffer cache and the rate at which redo is being generated
self adapting to sudden changes in workload
better performance
Turn on and off with innodb_adaptive_flushing

PAUSE in spin loops
Most people won’t care about this
But... if you are CPU bound - you will
Increases overall performance with CPU-bound workloads
Added benefit of minimizing power consumption during the execution of the spin loops.
Ugly details:
If the spin loops are executed too quickly, system resources are wasted, imposing a relatively severe penalty on transaction throughput
Uses a PAUSE instruction in its spin loops on all platforms where such an instruction is available.
Nothing needs to be done to take advantage of this

Faster Recovery
A number of optimizations speed up certain steps of the recovery that happens on the next startup after a crash.
Particularly scanning the redo log and applying the redo log are faster.
If you kept the size of your logfiles artificially low because recovery took a long time (1 hour to 1 day), you can consider increasing the logfile size.
No action needed to get this

Buffer Cache - Scan Resistant
Problem: with large scans it is possible to move all your “hot” data out of the buffer pool.
More IO loading the hot data back into the buffer pool
Slow response times
How to prevent it:
Mixed OLTP type with periodic batch reporting queries which result in large scans
set innodb_old_blocks_time for the duration of the batch run
Scanning large tables that cannot fit entirely in the buffer pool
set innodb_old_blocks_pct to a small value
Always benchmark to verify the effectiveness

Multiple Buffer Pools
Primarily useful for people with a large buffer pool size
Reduce the chances of encountering bottlenecks trying to access the same buffer pool
Pages are stored in or read from random buffer pools
Each buffer pool manages its own internal structures and is protected by its own buffer pool mutex
innodb_buffer_pool_instances
innodb_buffer_pool_size is divided up among all the buffer pools

Multiple Rollback Segments
History:
A bottleneck on high-capacity systems - currently it can handle a maximum of 1023 concurrent transactions that change any data
Limit on concurrent transactions is greatly expanded
Improves
Scalability - higher number of concurrent transactions
Performance - less contention when different threads access the rollback segments
To enable
Do a slow shutdown (innodb_fast_shutdown=0) before upgrading

Log Sys Mutex
Historically did double duty:
Controlled access to internal data structures related to log records and the LSN
Controlled access to pages in the buffer pool that are changed when a mini-transaction is committed.
Now broken out into separate mutexs
Improves performance
Nothing to enable - it is free

Questions?

http://www.mysqlpub.com	133
http://www.slideshare.net	31
http://phpprotip.com	20
http://www.planet-php.net	3
http://www.lmodules.com	1
http://mysqlpub.com	1

MySQL 5.5

by Ligaya Turmelle on May 20, 2010

Accessibility

Categories

Tags

Upload Details

Usage Rights

6 Embeds 189

Statistics

MySQL 5.5 — Presentation Transcript