SlideShare une entreprise Scribd logo
1  sur  81
Télécharger pour lire hors ligne
INNODB LOCKING
EXPLAINED
WITH STICK FIGURES
Bill Karwin
Bill Karwin
• Senior Database Architect at School Messenger
(West Corporation)
• Author of SQL Antipatterns: Avoiding the Pitfalls of
Database Programming
• 20+ years experience with SQL databases, software
development, MySQL consulting and training
why locking?
• When multiple clients
access the same data,
they have to avoid
clobbering each others’
work.
• Databases must restrict
access to one client at a
time for a given table or
row.
https://commons.wikimedia.org/wiki/File:New_York_City_Gridlock.jpg  
why locking?
• The DBMS creates locks
against tables and rows, and
gives them to clients, first-
come, first-serve.
• When a client requests a
exclusive lock, but a different
client currently holds it, the
requestor waits until the
holder releases its lock.
• Most locks last until the end of
the transaction.
https://commons.wikimedia.org/wiki/Traffic_lights#/media/File:LED_traffic_light.jpg
the analogy
a museum
• Many people can visit the museum (a database
table) to view art (rows of data).
reads and writes
reads
• Many visitors can view paintings at the same time—
no locking required.
https://commons.wikimedia.org/wiki/File:L%27%C3%A9glise_d%27Auvers-­‐‑sur-­‐‑Oise.jpg
writes
• A curator can change the paintings—while casual
visitors are viewing them.
https://commons.wikimedia.org/wiki/File:Vincent_Willem_van_Gogh_128.jpg
writes
• A curator can change the paintings—while casual
visitors are viewing them.
repeatable reads
• The viewers still see the prior painting in spite of the
change, because they used their tablets to capture
the image.*
*  Please  do  not  take  photographs  of  the  art  in  a  real  museum.
read committed
• If the viewers are okay allowing their view to
change, they can simply say so.
READ  
COMMITTED
read committed
• If the viewers are okay allowing their view to
change, they can simply say so.
READ  
COMMITTED
I  don’t  mind
exclusive locks
exclusive locks
• The curator can change the painting if they are the
exclusive person working on it.
exclusive locks
• Casual viewers don’t block the curator.
X-­‐‑lock!
exclusive locks
• A second curator who tries to change the same
painting must wait for the first curator to finish.
X-­‐‑lock!…
https://commons.wikimedia.org/wiki/Vincent_van_Gogh#/media/File:Vincent_Willem_van_Gogh_107.jpg
what is an exclusive lock?
• A lock that does not share.
• It must be the only lock on the resource.
• Request for an exclusive lock waits for the release of
any other shared or exclusive lock on that resource.
shared locks
shared locks
• The curator cannot make changes while the art
critic is viewing a painting.
S-­‐‑lock!
.  .  .
shared locks
• Art critics can share—they do not block each other.
S-­‐‑lock!
.  .  .
S-­‐‑lock!
shared locks
• Once the art critics leave, the curator can proceed
to change the art.
X-­‐‑lock!
shared locks
• A new art critic will not begin his viewing while the
curator is still working on changing the painting.
X-­‐‑lock!…
what is a shared lock?
• A lock that allows other shared locks on the same
table or rows.
• Shared locks blocks exclusive locks.
• Exclusive locks block shared locks.
table intention locks
table locks
• A construction worker needs to remodel the
museum, but not while visitors are inside.
table locks
• Each person is given a special badge as they enter
the museum, showing their
intention to view or change
the paintings.
IS!IX!
table locks
• A construction worker requests exclusive access to
the museum, but can not get it.
.  .  .IS!
IX!
table locks
• When all the visitors have left, the worker can finally
get his exclusive access to do his work.
LOCK  
TABLE!
table locks
• While the construction is going on, visitors cannot
get their badges, and therefore
cannot enter the museum.
…
https://commons.wikimedia.org/wiki/Category:Under_construction_icons#/media/File:Enobras.PNG
LOCK  
TABLE!
…
what is an intention lock?
• Before SELECT…LOCK IN SHARE MODE or other
shared lock, request an intention shared lock (“IS”)
on the table.
• Before SELECT…FOR UPDATE, INSERT, UPDATE,
DELETE, request an intention exclusive lock (“IX”).
• IS and IX locks allow access by multiple clients.
They won’t necessarily conflict until they try to get
real locks on the same rows.
• But a table lock (ALTER TABLE, DROP TABLE, LOCK
TABLES) blocks both IS and IX, and vice-versa.
gap locks
gap locks
• The art critic needs to view the whole collection
without changes and no new inserts.
gap
.  .  .
S-­‐‑lock!
what is a gap lock?
• A lock on a painting locks the “space” before the
painting.
• The gap lock prevents inserts of new paintings
before the locked painting (within the space).
• This happens automatically, to prevent “phantom
reads”—i.e. the view of data changes during a
transaction.
what is a gap lock?
• Exception: if the art critic is okay with seeing the
latest additions among the paintings, he can
choose to use the READ COMMITTED transaction
isolation level.
• Another exception: no gap lock is needed for a
UNIQUE or PRIMARY KEY index.
gap locks
• Optionally be more permissive about inserts.
gap
S-­‐‑lock!
READ  
COMMITTED
gap locks
• Optionally be more permissive about inserts.
X-­‐‑lock!
S-­‐‑lock!
READ  
COMMITTED
I  don’t  mind
traveling exhibit
• Some art exhibits move to another museum.
installation plans
• The curators must keep records of how to install the
art, so the show is the same at each museum.
Label  all  
paintings  on  
this  wall  with  
“Van  Gogh”
installation plans
• The curators must keep records of how to install the
art, so the show is the same at each museum.
Van  Gogh Van  Gogh
installation plans
• Another painting is added.
Van  Gogh Van  Gogh
Install  this  in  
the  gap  and  
label  it  
“Lautrec”
https://commons.wikimedia.org/wiki/Moulin_Rouge#/media/File:Lautrec_at_the_moulin_rouge_two_women_waltzing_1892.jpg
installation plans
• Another painting is installed in the gap.
Van  Gogh Van  GoghLautrec
installation plans
• The latter curator commits his installation plans first.
Van  Gogh Van  GoghLautrec
COMMIT
installation plans
• Then the first curator commits his installation plans.
Van  Gogh Van  GoghLautrec
COMMIT
traveling exhibit
• The exhibit is installed in the second museum.
order matters
• The installation instructions are carried out in the
wrong order!
Install  this  in  
the  gap  and  
label  it  
“Lautrec”
order matters
• The installation instructions are carried out in the
wrong order!
Lautrec
Label  all  
paintings  on  
this  wall  with  
“Van  Gogh”
order matters
• The installation instructions are carried out in the
wrong order!
Van  GoghVan  Gogh Van  Gogh
what happened?
• Installation instructions are recorded in the order
they were committed, not the order they were
originally done.
• Therefore the installation at the next museum may
repeat the steps in an incorrect order.
how to solve this?
• Labeling “Van Gogh” should have first locked all
the spaces on that wall.
Van  Gogh Van  Gogh
gap
how to solve this?
• The addition would be forced to wait for the locks
to be released.
Van  Gogh Van  Gogh
gap
…
insert intention locks
insert intention locks
• One curator wants to update paintings where
year > 1886
X-­‐‑lock!
gap
insert intention locks
• The second curator wants to insert an 1887 painting,
but it would fall within the existing gap lock.
X-­‐‑lock! …
gap
insert intention locks
---TRANSACTION 32070411, ACTIVE 6 sec inserting
mysql tables in use 1, locked 1
LOCK WAIT 2 lock struct(s), heap size 360, 1 row lock(s),
undo log entries 1
MySQL thread id 26, OS thread handle 0x7f2ba845f700,
query id 1423 192.168.50.1 root update
insert into Museum (year) values (1887)
------- TRX HAS BEEN WAITING 6 SEC FOR THIS LOCK TO BE
GRANTED:
RECORD LOCKS space id 3337 page no 4 n bits 72 index
`year` of table `test`.`Museum` trx id 32070411 lock_mode X
insert intention waiting
insert intention locks
• But if the first curator doesn’t care about new
paintings entering his view…
X-­‐‑lock!
gap
READ  
COMMITTED
insert intention locks
• Then the second curator gets an insert intention
lock, and is then free to insert.
insert-­‐‑
lock!
gap
X-­‐‑lock!
READ  
COMMITTED
I  don’t  mind
insert intention locks
• Then the second curator gets an insert intention
lock, and is then free to insert.
X-­‐‑lock!X-­‐‑lock!
READ  
COMMITTED
I  don’t  mind
what is an insert intention lock?
• A special kind of gap lock, requested before a
client tries to insert a row.
• Insert locks are shared, not exclusive—multiple
clients can acquire insert locks on the same gap.
• But insert locks conflict with other exclusive locks.
why is insert intention lock shared?
• Multiple clients prepare to insert into the same gap.
• They may be inserting different rows within the same
gap, so they don’t conflict with each other.
• But the insert intention lock blocks other clients from
requesting exclusive locks on the same gap.
auto-inc locks
auto-inc locks
• Two curators are installing paintings. They both need
to post a unique number for self-guided tours.
? ?
auto-inc locks
• There must be one number
generator per table.
• One curator at a time can
request the next value.
http://www.istockphoto.com/photo/ticket-­‐‑dispenser-­‐‑isolated-­‐‑9396862
auto-inc locks
• The first curator gets a number.
auto-­‐‑inc! 1 …
auto-inc locks
• As soon as the first curator gets his number, the
second curator can proceed.
1 2 auto-­‐‑inc!
auto-inc locks
• They may both keep locks on the paintings, but
they’re done allocating numbers.
X-­‐‑lock! 1 2 X-­‐‑lock!
what is an auto-inc lock?
• A table lock, used when a client requests the next
unique id for a given table.
• Ensures that each id is given to one client.
• Brief—it is released as soon as the id is generated,
instead of lasting to the end of the transaction like
other locks.
• Because the lock is so brief, neither client can
“undo”— i.e. they cannot return their id to the stack
for someone else to use.
deadlocks
deadlocks
• Two curators are updating the art, but they start
from different ends of the collection.
X-­‐‑lock!X-­‐‑lock!
deadlocks
• A curator requests a lock on the second painting,
which is already locked. He waits.
X-­‐‑lock!X-­‐‑lock!
…
deadlocks
• The other curator requests a lock on the first
painting, which is already locked. He also waits.
X-­‐‑lock!X-­‐‑lock!
…
…
deadlocks
• Neither will give up the lock they have, so they are
doomed to wait until one or both of them dies.
X-­‐‑lock!
X-­‐‑lock!
… …
what is a deadlock?
• When two or more concurrent clients wait for each
other to release their locks, but since they are both
waiting, they will never give up the lock they have.
• In other words, a “catch-22” of lock-waits.
• Many people use the term “deadlock” incorrectly—
when they are describing a simple one-way lock
wait.
resolving deadlocks
• MySQL detects cycles in lock waits, and kills one of
the transactions immediately.
X-­‐‑lock!X-­‐‑lock!
…
…
resolving deadlocks
• MySQL detects cycles in lock waits, and
automatically kills one of the transactions.
X-­‐‑lock!
X-­‐‑lock!
avoiding deadlocks
• All clients request locks in the same order.
X-­‐‑lock!
…
X-­‐‑lock!
avoiding deadlocks
• Each client locks everything they need in one
atomic request.
X-­‐‑lock!
…
conclusion
read committed
all the things?
• Yes … and no.
• READ COMMITTED avoids gap locks, therefore
reduces lock waits and allows greater throughput.
• But using READ COMMITTED allows more cases
where you could get deadlocks.
thank you
• I hope your trip to the museum was educational!
license and copyright
Copyright 2015-2016 Bill Karwin
http://www.slideshare.net/billkarwin
Released under a Creative Commons 3.0 License:
http://creativecommons.org/licenses/by-nc-nd/3.0/
You are free to share—to copy, distribute, and
transmit this work, under the following conditions:
Attribution.
You  must  attribute  this  
work  to  Bill  Karwin.
Noncommercial.
You  may  not  use  this  
work  for  commercial  
purposes.
No  Derivative  Works.
You may  not  alter,  
transform,  or  build  upon  
this  work.

Contenu connexe

Tendances

SeaweedFS introduction
SeaweedFS introductionSeaweedFS introduction
SeaweedFS introductionchrislusf
 
Multi Master PostgreSQL Cluster on Kubernetes
Multi Master PostgreSQL Cluster on KubernetesMulti Master PostgreSQL Cluster on Kubernetes
Multi Master PostgreSQL Cluster on KubernetesOhyama Masanori
 
My sql failover test using orchestrator
My sql failover test  using orchestratorMy sql failover test  using orchestrator
My sql failover test using orchestratorYoungHeon (Roy) Kim
 
All about Zookeeper and ClickHouse Keeper.pdf
All about Zookeeper and ClickHouse Keeper.pdfAll about Zookeeper and ClickHouse Keeper.pdf
All about Zookeeper and ClickHouse Keeper.pdfAltinity Ltd
 
Introduction to Redis
Introduction to RedisIntroduction to Redis
Introduction to RedisDvir Volk
 
Introduction and Overview of Apache Kafka, TriHUG July 23, 2013
Introduction and Overview of Apache Kafka, TriHUG July 23, 2013Introduction and Overview of Apache Kafka, TriHUG July 23, 2013
Introduction and Overview of Apache Kafka, TriHUG July 23, 2013mumrah
 
Solving PostgreSQL wicked problems
Solving PostgreSQL wicked problemsSolving PostgreSQL wicked problems
Solving PostgreSQL wicked problemsAlexander Korotkov
 
How to be Successful with Scylla
How to be Successful with ScyllaHow to be Successful with Scylla
How to be Successful with ScyllaScyllaDB
 
What Every Developer Should Know About Database Scalability
What Every Developer Should Know About Database ScalabilityWhat Every Developer Should Know About Database Scalability
What Every Developer Should Know About Database Scalabilityjbellis
 
MySQL Audit using Percona audit plugin and ELK
MySQL Audit using Percona audit plugin and ELKMySQL Audit using Percona audit plugin and ELK
MySQL Audit using Percona audit plugin and ELKYoungHeon (Roy) Kim
 
Advanced MySQL Query Tuning
Advanced MySQL Query TuningAdvanced MySQL Query Tuning
Advanced MySQL Query TuningAlexander Rubin
 
How to set up orchestrator to manage thousands of MySQL servers
How to set up orchestrator to manage thousands of MySQL serversHow to set up orchestrator to manage thousands of MySQL servers
How to set up orchestrator to manage thousands of MySQL serversSimon J Mudd
 
MMUG18 - MySQL Failover and Orchestrator
MMUG18 - MySQL Failover and OrchestratorMMUG18 - MySQL Failover and Orchestrator
MMUG18 - MySQL Failover and OrchestratorSimon J Mudd
 
MariaDB: in-depth (hands on training in Seoul)
MariaDB: in-depth (hands on training in Seoul)MariaDB: in-depth (hands on training in Seoul)
MariaDB: in-depth (hands on training in Seoul)Colin Charles
 
MySQL Performance Schema in Action
MySQL Performance Schema in ActionMySQL Performance Schema in Action
MySQL Performance Schema in ActionSveta Smirnova
 
Practical Partitioning in Production with Postgres
Practical Partitioning in Production with PostgresPractical Partitioning in Production with Postgres
Practical Partitioning in Production with PostgresEDB
 
My first 90 days with ClickHouse.pdf
My first 90 days with ClickHouse.pdfMy first 90 days with ClickHouse.pdf
My first 90 days with ClickHouse.pdfAlkin Tezuysal
 
Indexing with MongoDB
Indexing with MongoDBIndexing with MongoDB
Indexing with MongoDBMongoDB
 
patroni-based citrus high availability environment deployment
patroni-based citrus high availability environment deploymentpatroni-based citrus high availability environment deployment
patroni-based citrus high availability environment deploymenthyeongchae lee
 
MySQL Fabricでぼっこぼこにされたはなし
MySQL FabricでぼっこぼこにされたはなしMySQL Fabricでぼっこぼこにされたはなし
MySQL Fabricでぼっこぼこにされたはなしyoku0825
 

Tendances (20)

SeaweedFS introduction
SeaweedFS introductionSeaweedFS introduction
SeaweedFS introduction
 
Multi Master PostgreSQL Cluster on Kubernetes
Multi Master PostgreSQL Cluster on KubernetesMulti Master PostgreSQL Cluster on Kubernetes
Multi Master PostgreSQL Cluster on Kubernetes
 
My sql failover test using orchestrator
My sql failover test  using orchestratorMy sql failover test  using orchestrator
My sql failover test using orchestrator
 
All about Zookeeper and ClickHouse Keeper.pdf
All about Zookeeper and ClickHouse Keeper.pdfAll about Zookeeper and ClickHouse Keeper.pdf
All about Zookeeper and ClickHouse Keeper.pdf
 
Introduction to Redis
Introduction to RedisIntroduction to Redis
Introduction to Redis
 
Introduction and Overview of Apache Kafka, TriHUG July 23, 2013
Introduction and Overview of Apache Kafka, TriHUG July 23, 2013Introduction and Overview of Apache Kafka, TriHUG July 23, 2013
Introduction and Overview of Apache Kafka, TriHUG July 23, 2013
 
Solving PostgreSQL wicked problems
Solving PostgreSQL wicked problemsSolving PostgreSQL wicked problems
Solving PostgreSQL wicked problems
 
How to be Successful with Scylla
How to be Successful with ScyllaHow to be Successful with Scylla
How to be Successful with Scylla
 
What Every Developer Should Know About Database Scalability
What Every Developer Should Know About Database ScalabilityWhat Every Developer Should Know About Database Scalability
What Every Developer Should Know About Database Scalability
 
MySQL Audit using Percona audit plugin and ELK
MySQL Audit using Percona audit plugin and ELKMySQL Audit using Percona audit plugin and ELK
MySQL Audit using Percona audit plugin and ELK
 
Advanced MySQL Query Tuning
Advanced MySQL Query TuningAdvanced MySQL Query Tuning
Advanced MySQL Query Tuning
 
How to set up orchestrator to manage thousands of MySQL servers
How to set up orchestrator to manage thousands of MySQL serversHow to set up orchestrator to manage thousands of MySQL servers
How to set up orchestrator to manage thousands of MySQL servers
 
MMUG18 - MySQL Failover and Orchestrator
MMUG18 - MySQL Failover and OrchestratorMMUG18 - MySQL Failover and Orchestrator
MMUG18 - MySQL Failover and Orchestrator
 
MariaDB: in-depth (hands on training in Seoul)
MariaDB: in-depth (hands on training in Seoul)MariaDB: in-depth (hands on training in Seoul)
MariaDB: in-depth (hands on training in Seoul)
 
MySQL Performance Schema in Action
MySQL Performance Schema in ActionMySQL Performance Schema in Action
MySQL Performance Schema in Action
 
Practical Partitioning in Production with Postgres
Practical Partitioning in Production with PostgresPractical Partitioning in Production with Postgres
Practical Partitioning in Production with Postgres
 
My first 90 days with ClickHouse.pdf
My first 90 days with ClickHouse.pdfMy first 90 days with ClickHouse.pdf
My first 90 days with ClickHouse.pdf
 
Indexing with MongoDB
Indexing with MongoDBIndexing with MongoDB
Indexing with MongoDB
 
patroni-based citrus high availability environment deployment
patroni-based citrus high availability environment deploymentpatroni-based citrus high availability environment deployment
patroni-based citrus high availability environment deployment
 
MySQL Fabricでぼっこぼこにされたはなし
MySQL FabricでぼっこぼこにされたはなしMySQL Fabricでぼっこぼこにされたはなし
MySQL Fabricでぼっこぼこにされたはなし
 

En vedette

Citus Architecture: Extending Postgres to Build a Distributed Database
Citus Architecture: Extending Postgres to Build a Distributed DatabaseCitus Architecture: Extending Postgres to Build a Distributed Database
Citus Architecture: Extending Postgres to Build a Distributed DatabaseOzgun Erdogan
 
领域驱动设计与模型驱动开发
领域驱动设计与模型驱动开发领域驱动设计与模型驱动开发
领域驱动设计与模型驱动开发Weijun Zhong
 
Development with Qt for Windows CE
Development with Qt for Windows CEDevelopment with Qt for Windows CE
Development with Qt for Windows CEaccount inactive
 
python高级内存管理
python高级内存管理python高级内存管理
python高级内存管理rfyiamcool
 
Enterprise Architecture Implementation And The Open Group Architecture Framew...
Enterprise Architecture Implementation And The Open Group Architecture Framew...Enterprise Architecture Implementation And The Open Group Architecture Framew...
Enterprise Architecture Implementation And The Open Group Architecture Framew...Alan McSweeney
 
Mysql展示功能与源码对应
Mysql展示功能与源码对应Mysql展示功能与源码对应
Mysql展示功能与源码对应zhaolinjnu
 
MySQL High Availability Deep Dive
MySQL High Availability Deep DiveMySQL High Availability Deep Dive
MySQL High Availability Deep Divehastexo
 
Why MySQL Replication Fails, and How to Get it Back
Why MySQL Replication Fails, and How to Get it BackWhy MySQL Replication Fails, and How to Get it Back
Why MySQL Replication Fails, and How to Get it BackSveta Smirnova
 
MySQL Group Replication - HandsOn Tutorial
MySQL Group Replication - HandsOn TutorialMySQL Group Replication - HandsOn Tutorial
MySQL Group Replication - HandsOn TutorialKenny Gryp
 
MySQL Replication Performance Tuning for Fun and Profit!
MySQL Replication Performance Tuning for Fun and Profit!MySQL Replication Performance Tuning for Fun and Profit!
MySQL Replication Performance Tuning for Fun and Profit!Vitor Oliveira
 
Why MySQL High Availability Matters
Why MySQL High Availability MattersWhy MySQL High Availability Matters
Why MySQL High Availability MattersMatt Lord
 
Galera cluster for high availability
Galera cluster for high availability Galera cluster for high availability
Galera cluster for high availability Mydbops
 
Эффективная отладка репликации MySQL
Эффективная отладка репликации MySQLЭффективная отладка репликации MySQL
Эффективная отладка репликации MySQLSveta Smirnova
 
What you wanted to know about MySQL, but could not find using inernal instrum...
What you wanted to know about MySQL, but could not find using inernal instrum...What you wanted to know about MySQL, but could not find using inernal instrum...
What you wanted to know about MySQL, but could not find using inernal instrum...Sveta Smirnova
 
MySQL High Availability Solutions
MySQL High Availability SolutionsMySQL High Availability Solutions
MySQL High Availability SolutionsLenz Grimmer
 
Hbase源码初探
Hbase源码初探Hbase源码初探
Hbase源码初探zhaolinjnu
 
Мониторинг и отладка MySQL: максимум информации при минимальных потерях
Мониторинг и отладка MySQL: максимум информации при минимальных потеряхМониторинг и отладка MySQL: максимум информации при минимальных потерях
Мониторинг и отладка MySQL: максимум информации при минимальных потеряхSveta Smirnova
 
Advanced mysql replication techniques
Advanced mysql replication techniquesAdvanced mysql replication techniques
Advanced mysql replication techniquesGiuseppe Maxia
 

En vedette (20)

Citus Architecture: Extending Postgres to Build a Distributed Database
Citus Architecture: Extending Postgres to Build a Distributed DatabaseCitus Architecture: Extending Postgres to Build a Distributed Database
Citus Architecture: Extending Postgres to Build a Distributed Database
 
领域驱动设计与模型驱动开发
领域驱动设计与模型驱动开发领域驱动设计与模型驱动开发
领域驱动设计与模型驱动开发
 
Development with Qt for Windows CE
Development with Qt for Windows CEDevelopment with Qt for Windows CE
Development with Qt for Windows CE
 
python高级内存管理
python高级内存管理python高级内存管理
python高级内存管理
 
Enterprise Architecture Implementation And The Open Group Architecture Framew...
Enterprise Architecture Implementation And The Open Group Architecture Framew...Enterprise Architecture Implementation And The Open Group Architecture Framew...
Enterprise Architecture Implementation And The Open Group Architecture Framew...
 
Micro service
Micro serviceMicro service
Micro service
 
Mysql展示功能与源码对应
Mysql展示功能与源码对应Mysql展示功能与源码对应
Mysql展示功能与源码对应
 
Extensible Data Modeling
Extensible Data ModelingExtensible Data Modeling
Extensible Data Modeling
 
MySQL High Availability Deep Dive
MySQL High Availability Deep DiveMySQL High Availability Deep Dive
MySQL High Availability Deep Dive
 
Why MySQL Replication Fails, and How to Get it Back
Why MySQL Replication Fails, and How to Get it BackWhy MySQL Replication Fails, and How to Get it Back
Why MySQL Replication Fails, and How to Get it Back
 
MySQL Group Replication - HandsOn Tutorial
MySQL Group Replication - HandsOn TutorialMySQL Group Replication - HandsOn Tutorial
MySQL Group Replication - HandsOn Tutorial
 
MySQL Replication Performance Tuning for Fun and Profit!
MySQL Replication Performance Tuning for Fun and Profit!MySQL Replication Performance Tuning for Fun and Profit!
MySQL Replication Performance Tuning for Fun and Profit!
 
Why MySQL High Availability Matters
Why MySQL High Availability MattersWhy MySQL High Availability Matters
Why MySQL High Availability Matters
 
Galera cluster for high availability
Galera cluster for high availability Galera cluster for high availability
Galera cluster for high availability
 
Эффективная отладка репликации MySQL
Эффективная отладка репликации MySQLЭффективная отладка репликации MySQL
Эффективная отладка репликации MySQL
 
What you wanted to know about MySQL, but could not find using inernal instrum...
What you wanted to know about MySQL, but could not find using inernal instrum...What you wanted to know about MySQL, but could not find using inernal instrum...
What you wanted to know about MySQL, but could not find using inernal instrum...
 
MySQL High Availability Solutions
MySQL High Availability SolutionsMySQL High Availability Solutions
MySQL High Availability Solutions
 
Hbase源码初探
Hbase源码初探Hbase源码初探
Hbase源码初探
 
Мониторинг и отладка MySQL: максимум информации при минимальных потерях
Мониторинг и отладка MySQL: максимум информации при минимальных потеряхМониторинг и отладка MySQL: максимум информации при минимальных потерях
Мониторинг и отладка MySQL: максимум информации при минимальных потерях
 
Advanced mysql replication techniques
Advanced mysql replication techniquesAdvanced mysql replication techniques
Advanced mysql replication techniques
 

Plus de Karwin Software Solutions LLC (17)

How to Use JSON in MySQL Wrong
How to Use JSON in MySQL WrongHow to Use JSON in MySQL Wrong
How to Use JSON in MySQL Wrong
 
Recursive Query Throwdown
Recursive Query ThrowdownRecursive Query Throwdown
Recursive Query Throwdown
 
Load Data Fast!
Load Data Fast!Load Data Fast!
Load Data Fast!
 
SQL Outer Joins for Fun and Profit
SQL Outer Joins for Fun and ProfitSQL Outer Joins for Fun and Profit
SQL Outer Joins for Fun and Profit
 
Sql query patterns, optimized
Sql query patterns, optimizedSql query patterns, optimized
Sql query patterns, optimized
 
Survey of Percona Toolkit
Survey of Percona ToolkitSurvey of Percona Toolkit
Survey of Percona Toolkit
 
How to Design Indexes, Really
How to Design Indexes, ReallyHow to Design Indexes, Really
How to Design Indexes, Really
 
Schemadoc
SchemadocSchemadoc
Schemadoc
 
Percona toolkit
Percona toolkitPercona toolkit
Percona toolkit
 
MySQL 5.5 Guide to InnoDB Status
MySQL 5.5 Guide to InnoDB StatusMySQL 5.5 Guide to InnoDB Status
MySQL 5.5 Guide to InnoDB Status
 
Requirements the Last Bottleneck
Requirements the Last BottleneckRequirements the Last Bottleneck
Requirements the Last Bottleneck
 
Mentor Your Indexes
Mentor Your IndexesMentor Your Indexes
Mentor Your Indexes
 
Models for hierarchical data
Models for hierarchical dataModels for hierarchical data
Models for hierarchical data
 
Sql Injection Myths and Fallacies
Sql Injection Myths and FallaciesSql Injection Myths and Fallacies
Sql Injection Myths and Fallacies
 
Full Text Search In PostgreSQL
Full Text Search In PostgreSQLFull Text Search In PostgreSQL
Full Text Search In PostgreSQL
 
Practical Object Oriented Models In Sql
Practical Object Oriented Models In SqlPractical Object Oriented Models In Sql
Practical Object Oriented Models In Sql
 
Sql Antipatterns Strike Back
Sql Antipatterns Strike BackSql Antipatterns Strike Back
Sql Antipatterns Strike Back
 

Dernier

Search Engine Optimization SEO PDF for 2024.pdf
Search Engine Optimization SEO PDF for 2024.pdfSearch Engine Optimization SEO PDF for 2024.pdf
Search Engine Optimization SEO PDF for 2024.pdfRankYa
 
Anypoint Exchange: It’s Not Just a Repo!
Anypoint Exchange: It’s Not Just a Repo!Anypoint Exchange: It’s Not Just a Repo!
Anypoint Exchange: It’s Not Just a Repo!Manik S Magar
 
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr LapshynFwdays
 
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks..."LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...Fwdays
 
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...Patryk Bandurski
 
Connect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck PresentationConnect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck PresentationSlibray Presentation
 
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationBeyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationSafe Software
 
My Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 PresentationMy Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 PresentationRidwan Fadjar
 
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024BookNet Canada
 
Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?Mattias Andersson
 
SAP Build Work Zone - Overview L2-L3.pptx
SAP Build Work Zone - Overview L2-L3.pptxSAP Build Work Zone - Overview L2-L3.pptx
SAP Build Work Zone - Overview L2-L3.pptxNavinnSomaal
 
Powerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time ClashPowerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time Clashcharlottematthew16
 
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek SchlawackFwdays
 
Streamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupStreamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupFlorian Wilhelm
 
Commit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easyCommit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easyAlfredo García Lavilla
 
SIP trunking in Janus @ Kamailio World 2024
SIP trunking in Janus @ Kamailio World 2024SIP trunking in Janus @ Kamailio World 2024
SIP trunking in Janus @ Kamailio World 2024Lorenzo Miniero
 
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage CostLeverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage CostZilliz
 
Artificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptxArtificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptxhariprasad279825
 
Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Scott Keck-Warren
 

Dernier (20)

Search Engine Optimization SEO PDF for 2024.pdf
Search Engine Optimization SEO PDF for 2024.pdfSearch Engine Optimization SEO PDF for 2024.pdf
Search Engine Optimization SEO PDF for 2024.pdf
 
Anypoint Exchange: It’s Not Just a Repo!
Anypoint Exchange: It’s Not Just a Repo!Anypoint Exchange: It’s Not Just a Repo!
Anypoint Exchange: It’s Not Just a Repo!
 
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
 
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks..."LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
 
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
Integration and Automation in Practice: CI/CD in Mule Integration and Automat...
 
Connect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck PresentationConnect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck Presentation
 
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationBeyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
 
My Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 PresentationMy Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 Presentation
 
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
Transcript: New from BookNet Canada for 2024: BNC CataList - Tech Forum 2024
 
Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?
 
SAP Build Work Zone - Overview L2-L3.pptx
SAP Build Work Zone - Overview L2-L3.pptxSAP Build Work Zone - Overview L2-L3.pptx
SAP Build Work Zone - Overview L2-L3.pptx
 
Powerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time ClashPowerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time Clash
 
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
 
Streamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupStreamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project Setup
 
Commit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easyCommit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easy
 
SIP trunking in Janus @ Kamailio World 2024
SIP trunking in Janus @ Kamailio World 2024SIP trunking in Janus @ Kamailio World 2024
SIP trunking in Janus @ Kamailio World 2024
 
DMCC Future of Trade Web3 - Special Edition
DMCC Future of Trade Web3 - Special EditionDMCC Future of Trade Web3 - Special Edition
DMCC Future of Trade Web3 - Special Edition
 
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage CostLeverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
 
Artificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptxArtificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptx
 
Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024
 

InnoDB Locking Explained with Stick Figures

  • 2. Bill Karwin • Senior Database Architect at School Messenger (West Corporation) • Author of SQL Antipatterns: Avoiding the Pitfalls of Database Programming • 20+ years experience with SQL databases, software development, MySQL consulting and training
  • 3. why locking? • When multiple clients access the same data, they have to avoid clobbering each others’ work. • Databases must restrict access to one client at a time for a given table or row. https://commons.wikimedia.org/wiki/File:New_York_City_Gridlock.jpg  
  • 4. why locking? • The DBMS creates locks against tables and rows, and gives them to clients, first- come, first-serve. • When a client requests a exclusive lock, but a different client currently holds it, the requestor waits until the holder releases its lock. • Most locks last until the end of the transaction. https://commons.wikimedia.org/wiki/Traffic_lights#/media/File:LED_traffic_light.jpg
  • 6. a museum • Many people can visit the museum (a database table) to view art (rows of data).
  • 8. reads • Many visitors can view paintings at the same time— no locking required. https://commons.wikimedia.org/wiki/File:L%27%C3%A9glise_d%27Auvers-­‐‑sur-­‐‑Oise.jpg
  • 9. writes • A curator can change the paintings—while casual visitors are viewing them. https://commons.wikimedia.org/wiki/File:Vincent_Willem_van_Gogh_128.jpg
  • 10. writes • A curator can change the paintings—while casual visitors are viewing them.
  • 11. repeatable reads • The viewers still see the prior painting in spite of the change, because they used their tablets to capture the image.* *  Please  do  not  take  photographs  of  the  art  in  a  real  museum.
  • 12. read committed • If the viewers are okay allowing their view to change, they can simply say so. READ   COMMITTED
  • 13. read committed • If the viewers are okay allowing their view to change, they can simply say so. READ   COMMITTED I  don’t  mind
  • 15. exclusive locks • The curator can change the painting if they are the exclusive person working on it.
  • 16. exclusive locks • Casual viewers don’t block the curator. X-­‐‑lock!
  • 17. exclusive locks • A second curator who tries to change the same painting must wait for the first curator to finish. X-­‐‑lock!… https://commons.wikimedia.org/wiki/Vincent_van_Gogh#/media/File:Vincent_Willem_van_Gogh_107.jpg
  • 18. what is an exclusive lock? • A lock that does not share. • It must be the only lock on the resource. • Request for an exclusive lock waits for the release of any other shared or exclusive lock on that resource.
  • 20. shared locks • The curator cannot make changes while the art critic is viewing a painting. S-­‐‑lock! .  .  .
  • 21. shared locks • Art critics can share—they do not block each other. S-­‐‑lock! .  .  . S-­‐‑lock!
  • 22. shared locks • Once the art critics leave, the curator can proceed to change the art. X-­‐‑lock!
  • 23. shared locks • A new art critic will not begin his viewing while the curator is still working on changing the painting. X-­‐‑lock!…
  • 24. what is a shared lock? • A lock that allows other shared locks on the same table or rows. • Shared locks blocks exclusive locks. • Exclusive locks block shared locks.
  • 26. table locks • A construction worker needs to remodel the museum, but not while visitors are inside.
  • 27. table locks • Each person is given a special badge as they enter the museum, showing their intention to view or change the paintings. IS!IX!
  • 28. table locks • A construction worker requests exclusive access to the museum, but can not get it. .  .  .IS! IX!
  • 29. table locks • When all the visitors have left, the worker can finally get his exclusive access to do his work. LOCK   TABLE!
  • 30. table locks • While the construction is going on, visitors cannot get their badges, and therefore cannot enter the museum. … https://commons.wikimedia.org/wiki/Category:Under_construction_icons#/media/File:Enobras.PNG LOCK   TABLE! …
  • 31. what is an intention lock? • Before SELECT…LOCK IN SHARE MODE or other shared lock, request an intention shared lock (“IS”) on the table. • Before SELECT…FOR UPDATE, INSERT, UPDATE, DELETE, request an intention exclusive lock (“IX”). • IS and IX locks allow access by multiple clients. They won’t necessarily conflict until they try to get real locks on the same rows. • But a table lock (ALTER TABLE, DROP TABLE, LOCK TABLES) blocks both IS and IX, and vice-versa.
  • 33. gap locks • The art critic needs to view the whole collection without changes and no new inserts. gap .  .  . S-­‐‑lock!
  • 34. what is a gap lock? • A lock on a painting locks the “space” before the painting. • The gap lock prevents inserts of new paintings before the locked painting (within the space). • This happens automatically, to prevent “phantom reads”—i.e. the view of data changes during a transaction.
  • 35. what is a gap lock? • Exception: if the art critic is okay with seeing the latest additions among the paintings, he can choose to use the READ COMMITTED transaction isolation level. • Another exception: no gap lock is needed for a UNIQUE or PRIMARY KEY index.
  • 36. gap locks • Optionally be more permissive about inserts. gap S-­‐‑lock! READ   COMMITTED
  • 37. gap locks • Optionally be more permissive about inserts. X-­‐‑lock! S-­‐‑lock! READ   COMMITTED I  don’t  mind
  • 38. traveling exhibit • Some art exhibits move to another museum.
  • 39. installation plans • The curators must keep records of how to install the art, so the show is the same at each museum. Label  all   paintings  on   this  wall  with   “Van  Gogh”
  • 40. installation plans • The curators must keep records of how to install the art, so the show is the same at each museum. Van  Gogh Van  Gogh
  • 41. installation plans • Another painting is added. Van  Gogh Van  Gogh Install  this  in   the  gap  and   label  it   “Lautrec” https://commons.wikimedia.org/wiki/Moulin_Rouge#/media/File:Lautrec_at_the_moulin_rouge_two_women_waltzing_1892.jpg
  • 42. installation plans • Another painting is installed in the gap. Van  Gogh Van  GoghLautrec
  • 43. installation plans • The latter curator commits his installation plans first. Van  Gogh Van  GoghLautrec COMMIT
  • 44. installation plans • Then the first curator commits his installation plans. Van  Gogh Van  GoghLautrec COMMIT
  • 45. traveling exhibit • The exhibit is installed in the second museum.
  • 46. order matters • The installation instructions are carried out in the wrong order! Install  this  in   the  gap  and   label  it   “Lautrec”
  • 47. order matters • The installation instructions are carried out in the wrong order! Lautrec Label  all   paintings  on   this  wall  with   “Van  Gogh”
  • 48. order matters • The installation instructions are carried out in the wrong order! Van  GoghVan  Gogh Van  Gogh
  • 49. what happened? • Installation instructions are recorded in the order they were committed, not the order they were originally done. • Therefore the installation at the next museum may repeat the steps in an incorrect order.
  • 50. how to solve this? • Labeling “Van Gogh” should have first locked all the spaces on that wall. Van  Gogh Van  Gogh gap
  • 51. how to solve this? • The addition would be forced to wait for the locks to be released. Van  Gogh Van  Gogh gap …
  • 53. insert intention locks • One curator wants to update paintings where year > 1886 X-­‐‑lock! gap
  • 54. insert intention locks • The second curator wants to insert an 1887 painting, but it would fall within the existing gap lock. X-­‐‑lock! … gap
  • 55. insert intention locks ---TRANSACTION 32070411, ACTIVE 6 sec inserting mysql tables in use 1, locked 1 LOCK WAIT 2 lock struct(s), heap size 360, 1 row lock(s), undo log entries 1 MySQL thread id 26, OS thread handle 0x7f2ba845f700, query id 1423 192.168.50.1 root update insert into Museum (year) values (1887) ------- TRX HAS BEEN WAITING 6 SEC FOR THIS LOCK TO BE GRANTED: RECORD LOCKS space id 3337 page no 4 n bits 72 index `year` of table `test`.`Museum` trx id 32070411 lock_mode X insert intention waiting
  • 56. insert intention locks • But if the first curator doesn’t care about new paintings entering his view… X-­‐‑lock! gap READ   COMMITTED
  • 57. insert intention locks • Then the second curator gets an insert intention lock, and is then free to insert. insert-­‐‑ lock! gap X-­‐‑lock! READ   COMMITTED I  don’t  mind
  • 58. insert intention locks • Then the second curator gets an insert intention lock, and is then free to insert. X-­‐‑lock!X-­‐‑lock! READ   COMMITTED I  don’t  mind
  • 59. what is an insert intention lock? • A special kind of gap lock, requested before a client tries to insert a row. • Insert locks are shared, not exclusive—multiple clients can acquire insert locks on the same gap. • But insert locks conflict with other exclusive locks.
  • 60. why is insert intention lock shared? • Multiple clients prepare to insert into the same gap. • They may be inserting different rows within the same gap, so they don’t conflict with each other. • But the insert intention lock blocks other clients from requesting exclusive locks on the same gap.
  • 62. auto-inc locks • Two curators are installing paintings. They both need to post a unique number for self-guided tours. ? ?
  • 63. auto-inc locks • There must be one number generator per table. • One curator at a time can request the next value. http://www.istockphoto.com/photo/ticket-­‐‑dispenser-­‐‑isolated-­‐‑9396862
  • 64. auto-inc locks • The first curator gets a number. auto-­‐‑inc! 1 …
  • 65. auto-inc locks • As soon as the first curator gets his number, the second curator can proceed. 1 2 auto-­‐‑inc!
  • 66. auto-inc locks • They may both keep locks on the paintings, but they’re done allocating numbers. X-­‐‑lock! 1 2 X-­‐‑lock!
  • 67. what is an auto-inc lock? • A table lock, used when a client requests the next unique id for a given table. • Ensures that each id is given to one client. • Brief—it is released as soon as the id is generated, instead of lasting to the end of the transaction like other locks. • Because the lock is so brief, neither client can “undo”— i.e. they cannot return their id to the stack for someone else to use.
  • 69. deadlocks • Two curators are updating the art, but they start from different ends of the collection. X-­‐‑lock!X-­‐‑lock!
  • 70. deadlocks • A curator requests a lock on the second painting, which is already locked. He waits. X-­‐‑lock!X-­‐‑lock! …
  • 71. deadlocks • The other curator requests a lock on the first painting, which is already locked. He also waits. X-­‐‑lock!X-­‐‑lock! … …
  • 72. deadlocks • Neither will give up the lock they have, so they are doomed to wait until one or both of them dies. X-­‐‑lock! X-­‐‑lock! … …
  • 73. what is a deadlock? • When two or more concurrent clients wait for each other to release their locks, but since they are both waiting, they will never give up the lock they have. • In other words, a “catch-22” of lock-waits. • Many people use the term “deadlock” incorrectly— when they are describing a simple one-way lock wait.
  • 74. resolving deadlocks • MySQL detects cycles in lock waits, and kills one of the transactions immediately. X-­‐‑lock!X-­‐‑lock! … …
  • 75. resolving deadlocks • MySQL detects cycles in lock waits, and automatically kills one of the transactions. X-­‐‑lock! X-­‐‑lock!
  • 76. avoiding deadlocks • All clients request locks in the same order. X-­‐‑lock! … X-­‐‑lock!
  • 77. avoiding deadlocks • Each client locks everything they need in one atomic request. X-­‐‑lock! …
  • 79. read committed all the things? • Yes … and no. • READ COMMITTED avoids gap locks, therefore reduces lock waits and allows greater throughput. • But using READ COMMITTED allows more cases where you could get deadlocks.
  • 80. thank you • I hope your trip to the museum was educational!
  • 81. license and copyright Copyright 2015-2016 Bill Karwin http://www.slideshare.net/billkarwin Released under a Creative Commons 3.0 License: http://creativecommons.org/licenses/by-nc-nd/3.0/ You are free to share—to copy, distribute, and transmit this work, under the following conditions: Attribution. You  must  attribute  this   work  to  Bill  Karwin. Noncommercial. You  may  not  use  this   work  for  commercial   purposes. No  Derivative  Works. You may  not  alter,   transform,  or  build  upon   this  work.