SlideShare a Scribd company logo
SPANNER
WS2019/2020
Google's Globally-Distributed
Database
BDASEM
Ahmed Amine Mchayaa
Technische Universität
Berlin
Gabor E. Gevay
Supervised by :
Outline
Intro :
What is
Spanner
Architectur
e behind
Spanner
Spanner:
a CA
system
Summary
2
Annual revenue of Google from 2005 to 2018 (in
Billion U.S dollars)
Revenue in billion
U.S. dollars
6.1
Sources: Google; Statista 2019
2005
29.3
2010
74.54
2015
136.22
2018
3
Background
Horizontally
Scaling Database
ADWARDS
ACID
Transactions with
global
consistency
No DownTime
Google needs
4
Globally
distributed
Fully managed,
database service
with global scale
Traditional
relational
semantics:
Schemas, ACID
transaction, SQL
Semi -Relational
Database
Synchronously
replicated
Automatic,
synchronous
replication within
and across regions
for availability
What is Spanner?
5
Architecture OverViewComputeStorage
DB 1
DB n
DB 1
DB n
DB 1
DB n
Zone 1 Zone 2 Zone 3
Regional Instance
6
Sources: Robert K.
Spanner - a fully managed horizontally
scalable relational database...
Architecture Overview
DB 1
DB 2
DB 3
DB 4
DB 5
DB n
Instance
Split 1
Split 2
Split 3
Split 4
Split 5
Split 6
Split 7
Split n
Table 1
Table 2
Table 3
Table 4
Table 5
Table 6
Table 7
Table n
Zone
7
Sources: Robert K.
Spanner - a fully managed horizontally
scalable relational database...
Architecture OverView
Split 1
Zone 1
ComputeStorage
Split 2
Split 3
Paxos
Group
for Split
1
Split 1
Zone 1
Split 2
Split 3
Split 1
Zone 1
Split 2
Split 3
*TrueTime used for leader leases : There is only one leader for a split at any given time
Regional Instance
8
Sources: Robert K.
Spanner - a fully managed horizontally
scalable relational database...
• Invented during the creation of Spanner
• Quantifies the « worst » possible error / drift between clocks in all
datacenters arount the world (global clock)
• TrueTime.now() gives you an interval [t1,t2]; t2 = t1+ 2Є
• t1 is guaranteed to be lower than the value of the global clock at
the instant when Now() finishes executing
• T2 is guaranteed to be higher than the value of the global clock at
the instant when Now() starts executing
9
Architecture OverView – True Time
10
Architecture OverView – True Time
Computer NodeComputer NodeComputer Node
Atomic MasterAtomic Master Atomic Master
GPS Master GPS Master GPS Master
Sync every 30 sec
Sync every 30 sec, Synchonization within 50µs, ε guaranteed interval around 2ms
Life of query : Consistent Read
Split 1
Zone 1
Split 2
Split 3
Split 1
Zone 1
Split 2
Split 3
Split 1
Zone 1
Split 2
Split 3
11
Sources: Robert K.
Spanner - a fully managed horizontally
scalable relational database...
slave leader slave
4. Wait for data /
Response
1. Request
3 No/ Yes
2. Ok
to read
Life of query : Stale Read
Split 1
Zone 1
Split 2
Split 3
Split 1
Zone 1
Split 2
Split 3
Split 1
Zone 1
Split 2
Split 3
12
Sources: Robert K.
Spanner - a fully managed horizontally
scalable relational database...
slave leader slave
4. Wait for data /
Response
1. Request
(max 15s old )
2. Am I up-
to-date
Life of query : Read
Split 1
Zone 1
Split 2
Split 3
Split 1
Zone 1
Split 2
Split 3
Split 1
Zone 1
Split 2
Split 3
13
Sources: Robert K.
Spanner - a fully managed horizontally
scalable relational database...
slave
leader slave
3. Query result
4. txn. Bufferwrite
1. txn. Query()
2. acq. locks
5. Write 5. Write
6. ack6. ack 7. Rel. locks
14
Spanner claims to be consistent and available
CA CP
AP
It is impossible for a
distributed computer
system to simultaneously
provide more than two
out of three of the
following guarantees:
Consistency, Availability,
Partition Tolerance
15
Always able
to read and
Write
Always see the
same data as
others at same
point in time
Works even
in the case
of network
partition
Pick Two !!!
Summary
Read / Write Transactions
What Google Spanner is,
the idea behind and and
what offers as database
The Architecture behind
Google Spanner
16
Spanner a CA/CP
system
David F. Bacon et al. 2017. Spanner: Becoming a SQL System.
In Proceedings of the 2017 ACM International Conference on Management of Data (SIGMOD '17). ACM,
New York, NY, USA, 331-343. DOI: https://doi.org/10.1145/3035918.3056103
1
James C. Corbett et al. 2012. Spanner: Google's globally-distributed database
In Proceedings of the 10th USENIX conference on Operating Systems Design and Implementation
(OSDI'12). USENIX Association, Berkeley, CA, USA, 251-264.
2
Jeff Shute et al. 2013. F1: a distributed SQL database that scales.
Proc. VLDB Endow. 6, 11 (August 2013), 1068-1079. DOI: http://dx.doi.org/10.14778/2536222.2536232
3
Eric B. 2017. Spanner, TrueTime & The CAP Theorem
VP, Infrastructure, Google. February 14, 2017 :https://ai.google/research/pubs/pub45855
4
Robert K. 2017. Spanner - a fully managed horizontally scalable relational database
DEVOXX Poland (June 2017) [Video], https://www.youtube.com/watch?v=IFbydfGV2lQ
5
References
17

More Related Content

Similar to Spanner : Google' s Globally Distributed Database

What’s New in ScyllaDB Open Source 5.0
What’s New in ScyllaDB Open Source 5.0What’s New in ScyllaDB Open Source 5.0
What’s New in ScyllaDB Open Source 5.0
ScyllaDB
 
From 1000/day to 1000/sec: The Evolution of Incapsula's BIG DATA System [Surg...
From 1000/day to 1000/sec: The Evolution of Incapsula's BIG DATA System [Surg...From 1000/day to 1000/sec: The Evolution of Incapsula's BIG DATA System [Surg...
From 1000/day to 1000/sec: The Evolution of Incapsula's BIG DATA System [Surg...
Imperva Incapsula
 
ScyllaDB V Developer Deep Dive Series: Resiliency and Strong Consistency via ...
ScyllaDB V Developer Deep Dive Series: Resiliency and Strong Consistency via ...ScyllaDB V Developer Deep Dive Series: Resiliency and Strong Consistency via ...
ScyllaDB V Developer Deep Dive Series: Resiliency and Strong Consistency via ...
ScyllaDB
 
Cloudstack Open source and you
Cloudstack Open source and you Cloudstack Open source and you
Cloudstack Open source and you Brian Bullard
 
Azure Cosmos DB - NoSQL Strikes Back (An introduction to the dark side of you...
Azure Cosmos DB - NoSQL Strikes Back (An introduction to the dark side of you...Azure Cosmos DB - NoSQL Strikes Back (An introduction to the dark side of you...
Azure Cosmos DB - NoSQL Strikes Back (An introduction to the dark side of you...
Andre Essing
 
[PASS Summit 2016] Blazing Fast, Planet-Scale Customer Scenarios with Azure D...
[PASS Summit 2016] Blazing Fast, Planet-Scale Customer Scenarios with Azure D...[PASS Summit 2016] Blazing Fast, Planet-Scale Customer Scenarios with Azure D...
[PASS Summit 2016] Blazing Fast, Planet-Scale Customer Scenarios with Azure D...
Andrew Liu
 
Azure Cosmos DB - Technical Deep Dive
Azure Cosmos DB - Technical Deep DiveAzure Cosmos DB - Technical Deep Dive
Azure Cosmos DB - Technical Deep Dive
Andre Essing
 
Google Cloud Spanner y NewSQL
Google Cloud Spanner y NewSQLGoogle Cloud Spanner y NewSQL
Google Cloud Spanner y NewSQL
Globant
 
OpenStack at Scale Inside NetApp
OpenStack at Scale Inside NetAppOpenStack at Scale Inside NetApp
OpenStack at Scale Inside NetApp
Tesora
 
Closing Keynote
Closing KeynoteClosing Keynote
Closing Keynote
Neo4j
 
Mysql Latency
Mysql LatencyMysql Latency
Mysql Latency
srubinstein
 
Apache Beam (incubating)
Apache Beam (incubating)Apache Beam (incubating)
Apache Beam (incubating)
Apache Apex
 
Spark Summit EU talk by Sameer Agarwal
Spark Summit EU talk by Sameer AgarwalSpark Summit EU talk by Sameer Agarwal
Spark Summit EU talk by Sameer Agarwal
Spark Summit
 
Management and Automation of MongoDB Clusters - Slides
Management and Automation of MongoDB Clusters - SlidesManagement and Automation of MongoDB Clusters - Slides
Management and Automation of MongoDB Clusters - Slides
Severalnines
 
Flink at netflix paypal speaker series
Flink at netflix   paypal speaker seriesFlink at netflix   paypal speaker series
Flink at netflix paypal speaker series
Monal Daxini
 
Apache Beam: A unified model for batch and stream processing data
Apache Beam: A unified model for batch and stream processing dataApache Beam: A unified model for batch and stream processing data
Apache Beam: A unified model for batch and stream processing data
DataWorks Summit/Hadoop Summit
 
101 ways to configure kafka - badly
101 ways to configure kafka - badly101 ways to configure kafka - badly
101 ways to configure kafka - badly
Henning Spjelkavik
 
Strata Conference + Hadoop World NY 2016: Lessons learned building a scalable...
Strata Conference + Hadoop World NY 2016: Lessons learned building a scalable...Strata Conference + Hadoop World NY 2016: Lessons learned building a scalable...
Strata Conference + Hadoop World NY 2016: Lessons learned building a scalable...
Sumeet Singh
 
MongoDB World 2018: Solving Your Backup Needs Using MongoDB Ops Manager, Clou...
MongoDB World 2018: Solving Your Backup Needs Using MongoDB Ops Manager, Clou...MongoDB World 2018: Solving Your Backup Needs Using MongoDB Ops Manager, Clou...
MongoDB World 2018: Solving Your Backup Needs Using MongoDB Ops Manager, Clou...
MongoDB
 

Similar to Spanner : Google' s Globally Distributed Database (20)

What’s New in ScyllaDB Open Source 5.0
What’s New in ScyllaDB Open Source 5.0What’s New in ScyllaDB Open Source 5.0
What’s New in ScyllaDB Open Source 5.0
 
From 1000/day to 1000/sec: The Evolution of Incapsula's BIG DATA System [Surg...
From 1000/day to 1000/sec: The Evolution of Incapsula's BIG DATA System [Surg...From 1000/day to 1000/sec: The Evolution of Incapsula's BIG DATA System [Surg...
From 1000/day to 1000/sec: The Evolution of Incapsula's BIG DATA System [Surg...
 
ScyllaDB V Developer Deep Dive Series: Resiliency and Strong Consistency via ...
ScyllaDB V Developer Deep Dive Series: Resiliency and Strong Consistency via ...ScyllaDB V Developer Deep Dive Series: Resiliency and Strong Consistency via ...
ScyllaDB V Developer Deep Dive Series: Resiliency and Strong Consistency via ...
 
Cloudstack Open source and you
Cloudstack Open source and you Cloudstack Open source and you
Cloudstack Open source and you
 
Azure Cosmos DB - NoSQL Strikes Back (An introduction to the dark side of you...
Azure Cosmos DB - NoSQL Strikes Back (An introduction to the dark side of you...Azure Cosmos DB - NoSQL Strikes Back (An introduction to the dark side of you...
Azure Cosmos DB - NoSQL Strikes Back (An introduction to the dark side of you...
 
[PASS Summit 2016] Blazing Fast, Planet-Scale Customer Scenarios with Azure D...
[PASS Summit 2016] Blazing Fast, Planet-Scale Customer Scenarios with Azure D...[PASS Summit 2016] Blazing Fast, Planet-Scale Customer Scenarios with Azure D...
[PASS Summit 2016] Blazing Fast, Planet-Scale Customer Scenarios with Azure D...
 
Azure Cosmos DB - Technical Deep Dive
Azure Cosmos DB - Technical Deep DiveAzure Cosmos DB - Technical Deep Dive
Azure Cosmos DB - Technical Deep Dive
 
Google Cloud Spanner y NewSQL
Google Cloud Spanner y NewSQLGoogle Cloud Spanner y NewSQL
Google Cloud Spanner y NewSQL
 
OpenStack at Scale Inside NetApp
OpenStack at Scale Inside NetAppOpenStack at Scale Inside NetApp
OpenStack at Scale Inside NetApp
 
Closing Keynote
Closing KeynoteClosing Keynote
Closing Keynote
 
Mysql Latency
Mysql LatencyMysql Latency
Mysql Latency
 
Apache Beam (incubating)
Apache Beam (incubating)Apache Beam (incubating)
Apache Beam (incubating)
 
Spark Summit EU talk by Sameer Agarwal
Spark Summit EU talk by Sameer AgarwalSpark Summit EU talk by Sameer Agarwal
Spark Summit EU talk by Sameer Agarwal
 
Management and Automation of MongoDB Clusters - Slides
Management and Automation of MongoDB Clusters - SlidesManagement and Automation of MongoDB Clusters - Slides
Management and Automation of MongoDB Clusters - Slides
 
Flink at netflix paypal speaker series
Flink at netflix   paypal speaker seriesFlink at netflix   paypal speaker series
Flink at netflix paypal speaker series
 
Handout3o
Handout3oHandout3o
Handout3o
 
Apache Beam: A unified model for batch and stream processing data
Apache Beam: A unified model for batch and stream processing dataApache Beam: A unified model for batch and stream processing data
Apache Beam: A unified model for batch and stream processing data
 
101 ways to configure kafka - badly
101 ways to configure kafka - badly101 ways to configure kafka - badly
101 ways to configure kafka - badly
 
Strata Conference + Hadoop World NY 2016: Lessons learned building a scalable...
Strata Conference + Hadoop World NY 2016: Lessons learned building a scalable...Strata Conference + Hadoop World NY 2016: Lessons learned building a scalable...
Strata Conference + Hadoop World NY 2016: Lessons learned building a scalable...
 
MongoDB World 2018: Solving Your Backup Needs Using MongoDB Ops Manager, Clou...
MongoDB World 2018: Solving Your Backup Needs Using MongoDB Ops Manager, Clou...MongoDB World 2018: Solving Your Backup Needs Using MongoDB Ops Manager, Clou...
MongoDB World 2018: Solving Your Backup Needs Using MongoDB Ops Manager, Clou...
 

Recently uploaded

一比一原版(QU毕业证)皇后大学毕业证成绩单
一比一原版(QU毕业证)皇后大学毕业证成绩单一比一原版(QU毕业证)皇后大学毕业证成绩单
一比一原版(QU毕业证)皇后大学毕业证成绩单
enxupq
 
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
John Andrews
 
Ch03-Managing the Object-Oriented Information Systems Project a.pdf
Ch03-Managing the Object-Oriented Information Systems Project a.pdfCh03-Managing the Object-Oriented Information Systems Project a.pdf
Ch03-Managing the Object-Oriented Information Systems Project a.pdf
haila53
 
Innovative Methods in Media and Communication Research by Sebastian Kubitschk...
Innovative Methods in Media and Communication Research by Sebastian Kubitschk...Innovative Methods in Media and Communication Research by Sebastian Kubitschk...
Innovative Methods in Media and Communication Research by Sebastian Kubitschk...
correoyaya
 
一比一原版(CBU毕业证)不列颠海角大学毕业证成绩单
一比一原版(CBU毕业证)不列颠海角大学毕业证成绩单一比一原版(CBU毕业证)不列颠海角大学毕业证成绩单
一比一原版(CBU毕业证)不列颠海角大学毕业证成绩单
nscud
 
The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...
jerlynmaetalle
 
做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样
做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样
做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样
axoqas
 
Q1’2024 Update: MYCI’s Leap Year Rebound
Q1’2024 Update: MYCI’s Leap Year ReboundQ1’2024 Update: MYCI’s Leap Year Rebound
Q1’2024 Update: MYCI’s Leap Year Rebound
Oppotus
 
1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx
1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx
1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx
Tiktokethiodaily
 
standardisation of garbhpala offhgfffghh
standardisation of garbhpala offhgfffghhstandardisation of garbhpala offhgfffghh
standardisation of garbhpala offhgfffghh
ArpitMalhotra16
 
一比一原版(NYU毕业证)纽约大学毕业证成绩单
一比一原版(NYU毕业证)纽约大学毕业证成绩单一比一原版(NYU毕业证)纽约大学毕业证成绩单
一比一原版(NYU毕业证)纽约大学毕业证成绩单
ewymefz
 
Opendatabay - Open Data Marketplace.pptx
Opendatabay - Open Data Marketplace.pptxOpendatabay - Open Data Marketplace.pptx
Opendatabay - Open Data Marketplace.pptx
Opendatabay
 
Jpolillo Amazon PPC - Bid Optimization Sample
Jpolillo Amazon PPC - Bid Optimization SampleJpolillo Amazon PPC - Bid Optimization Sample
Jpolillo Amazon PPC - Bid Optimization Sample
James Polillo
 
一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单
一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单
一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单
ewymefz
 
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Subhajit Sahu
 
一比一原版(TWU毕业证)西三一大学毕业证成绩单
一比一原版(TWU毕业证)西三一大学毕业证成绩单一比一原版(TWU毕业证)西三一大学毕业证成绩单
一比一原版(TWU毕业证)西三一大学毕业证成绩单
ocavb
 
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
AbhimanyuSinha9
 
SOCRadar Germany 2024 Threat Landscape Report
SOCRadar Germany 2024 Threat Landscape ReportSOCRadar Germany 2024 Threat Landscape Report
SOCRadar Germany 2024 Threat Landscape Report
SOCRadar
 
Empowering Data Analytics Ecosystem.pptx
Empowering Data Analytics Ecosystem.pptxEmpowering Data Analytics Ecosystem.pptx
Empowering Data Analytics Ecosystem.pptx
benishzehra469
 
社内勉強会資料_LLM Agents                              .
社内勉強会資料_LLM Agents                              .社内勉強会資料_LLM Agents                              .
社内勉強会資料_LLM Agents                              .
NABLAS株式会社
 

Recently uploaded (20)

一比一原版(QU毕业证)皇后大学毕业证成绩单
一比一原版(QU毕业证)皇后大学毕业证成绩单一比一原版(QU毕业证)皇后大学毕业证成绩单
一比一原版(QU毕业证)皇后大学毕业证成绩单
 
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...
 
Ch03-Managing the Object-Oriented Information Systems Project a.pdf
Ch03-Managing the Object-Oriented Information Systems Project a.pdfCh03-Managing the Object-Oriented Information Systems Project a.pdf
Ch03-Managing the Object-Oriented Information Systems Project a.pdf
 
Innovative Methods in Media and Communication Research by Sebastian Kubitschk...
Innovative Methods in Media and Communication Research by Sebastian Kubitschk...Innovative Methods in Media and Communication Research by Sebastian Kubitschk...
Innovative Methods in Media and Communication Research by Sebastian Kubitschk...
 
一比一原版(CBU毕业证)不列颠海角大学毕业证成绩单
一比一原版(CBU毕业证)不列颠海角大学毕业证成绩单一比一原版(CBU毕业证)不列颠海角大学毕业证成绩单
一比一原版(CBU毕业证)不列颠海角大学毕业证成绩单
 
The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...The affect of service quality and online reviews on customer loyalty in the E...
The affect of service quality and online reviews on customer loyalty in the E...
 
做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样
做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样
做(mqu毕业证书)麦考瑞大学毕业证硕士文凭证书学费发票原版一模一样
 
Q1’2024 Update: MYCI’s Leap Year Rebound
Q1’2024 Update: MYCI’s Leap Year ReboundQ1’2024 Update: MYCI’s Leap Year Rebound
Q1’2024 Update: MYCI’s Leap Year Rebound
 
1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx
1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx
1.Seydhcuxhxyxhccuuxuxyxyxmisolids 2019.pptx
 
standardisation of garbhpala offhgfffghh
standardisation of garbhpala offhgfffghhstandardisation of garbhpala offhgfffghh
standardisation of garbhpala offhgfffghh
 
一比一原版(NYU毕业证)纽约大学毕业证成绩单
一比一原版(NYU毕业证)纽约大学毕业证成绩单一比一原版(NYU毕业证)纽约大学毕业证成绩单
一比一原版(NYU毕业证)纽约大学毕业证成绩单
 
Opendatabay - Open Data Marketplace.pptx
Opendatabay - Open Data Marketplace.pptxOpendatabay - Open Data Marketplace.pptx
Opendatabay - Open Data Marketplace.pptx
 
Jpolillo Amazon PPC - Bid Optimization Sample
Jpolillo Amazon PPC - Bid Optimization SampleJpolillo Amazon PPC - Bid Optimization Sample
Jpolillo Amazon PPC - Bid Optimization Sample
 
一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单
一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单
一比一原版(UMich毕业证)密歇根大学|安娜堡分校毕业证成绩单
 
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...
 
一比一原版(TWU毕业证)西三一大学毕业证成绩单
一比一原版(TWU毕业证)西三一大学毕业证成绩单一比一原版(TWU毕业证)西三一大学毕业证成绩单
一比一原版(TWU毕业证)西三一大学毕业证成绩单
 
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...Best best suvichar in gujarati english meaning of this sentence as Silk road ...
Best best suvichar in gujarati english meaning of this sentence as Silk road ...
 
SOCRadar Germany 2024 Threat Landscape Report
SOCRadar Germany 2024 Threat Landscape ReportSOCRadar Germany 2024 Threat Landscape Report
SOCRadar Germany 2024 Threat Landscape Report
 
Empowering Data Analytics Ecosystem.pptx
Empowering Data Analytics Ecosystem.pptxEmpowering Data Analytics Ecosystem.pptx
Empowering Data Analytics Ecosystem.pptx
 
社内勉強会資料_LLM Agents                              .
社内勉強会資料_LLM Agents                              .社内勉強会資料_LLM Agents                              .
社内勉強会資料_LLM Agents                              .
 

Spanner : Google' s Globally Distributed Database

  • 1. SPANNER WS2019/2020 Google's Globally-Distributed Database BDASEM Ahmed Amine Mchayaa Technische Universität Berlin Gabor E. Gevay Supervised by :
  • 2. Outline Intro : What is Spanner Architectur e behind Spanner Spanner: a CA system Summary 2
  • 3. Annual revenue of Google from 2005 to 2018 (in Billion U.S dollars) Revenue in billion U.S. dollars 6.1 Sources: Google; Statista 2019 2005 29.3 2010 74.54 2015 136.22 2018 3 Background
  • 5. Globally distributed Fully managed, database service with global scale Traditional relational semantics: Schemas, ACID transaction, SQL Semi -Relational Database Synchronously replicated Automatic, synchronous replication within and across regions for availability What is Spanner? 5
  • 6. Architecture OverViewComputeStorage DB 1 DB n DB 1 DB n DB 1 DB n Zone 1 Zone 2 Zone 3 Regional Instance 6 Sources: Robert K. Spanner - a fully managed horizontally scalable relational database...
  • 7. Architecture Overview DB 1 DB 2 DB 3 DB 4 DB 5 DB n Instance Split 1 Split 2 Split 3 Split 4 Split 5 Split 6 Split 7 Split n Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table n Zone 7 Sources: Robert K. Spanner - a fully managed horizontally scalable relational database...
  • 8. Architecture OverView Split 1 Zone 1 ComputeStorage Split 2 Split 3 Paxos Group for Split 1 Split 1 Zone 1 Split 2 Split 3 Split 1 Zone 1 Split 2 Split 3 *TrueTime used for leader leases : There is only one leader for a split at any given time Regional Instance 8 Sources: Robert K. Spanner - a fully managed horizontally scalable relational database...
  • 9. • Invented during the creation of Spanner • Quantifies the « worst » possible error / drift between clocks in all datacenters arount the world (global clock) • TrueTime.now() gives you an interval [t1,t2]; t2 = t1+ 2Є • t1 is guaranteed to be lower than the value of the global clock at the instant when Now() finishes executing • T2 is guaranteed to be higher than the value of the global clock at the instant when Now() starts executing 9 Architecture OverView – True Time
  • 10. 10 Architecture OverView – True Time Computer NodeComputer NodeComputer Node Atomic MasterAtomic Master Atomic Master GPS Master GPS Master GPS Master Sync every 30 sec Sync every 30 sec, Synchonization within 50µs, ε guaranteed interval around 2ms
  • 11. Life of query : Consistent Read Split 1 Zone 1 Split 2 Split 3 Split 1 Zone 1 Split 2 Split 3 Split 1 Zone 1 Split 2 Split 3 11 Sources: Robert K. Spanner - a fully managed horizontally scalable relational database... slave leader slave 4. Wait for data / Response 1. Request 3 No/ Yes 2. Ok to read
  • 12. Life of query : Stale Read Split 1 Zone 1 Split 2 Split 3 Split 1 Zone 1 Split 2 Split 3 Split 1 Zone 1 Split 2 Split 3 12 Sources: Robert K. Spanner - a fully managed horizontally scalable relational database... slave leader slave 4. Wait for data / Response 1. Request (max 15s old ) 2. Am I up- to-date
  • 13. Life of query : Read Split 1 Zone 1 Split 2 Split 3 Split 1 Zone 1 Split 2 Split 3 Split 1 Zone 1 Split 2 Split 3 13 Sources: Robert K. Spanner - a fully managed horizontally scalable relational database... slave leader slave 3. Query result 4. txn. Bufferwrite 1. txn. Query() 2. acq. locks 5. Write 5. Write 6. ack6. ack 7. Rel. locks
  • 14. 14
  • 15. Spanner claims to be consistent and available CA CP AP It is impossible for a distributed computer system to simultaneously provide more than two out of three of the following guarantees: Consistency, Availability, Partition Tolerance 15 Always able to read and Write Always see the same data as others at same point in time Works even in the case of network partition Pick Two !!!
  • 16. Summary Read / Write Transactions What Google Spanner is, the idea behind and and what offers as database The Architecture behind Google Spanner 16 Spanner a CA/CP system
  • 17. David F. Bacon et al. 2017. Spanner: Becoming a SQL System. In Proceedings of the 2017 ACM International Conference on Management of Data (SIGMOD '17). ACM, New York, NY, USA, 331-343. DOI: https://doi.org/10.1145/3035918.3056103 1 James C. Corbett et al. 2012. Spanner: Google's globally-distributed database In Proceedings of the 10th USENIX conference on Operating Systems Design and Implementation (OSDI'12). USENIX Association, Berkeley, CA, USA, 251-264. 2 Jeff Shute et al. 2013. F1: a distributed SQL database that scales. Proc. VLDB Endow. 6, 11 (August 2013), 1068-1079. DOI: http://dx.doi.org/10.14778/2536222.2536232 3 Eric B. 2017. Spanner, TrueTime & The CAP Theorem VP, Infrastructure, Google. February 14, 2017 :https://ai.google/research/pubs/pub45855 4 Robert K. 2017. Spanner - a fully managed horizontally scalable relational database DEVOXX Poland (June 2017) [Video], https://www.youtube.com/watch?v=IFbydfGV2lQ 5 References 17