SlideShare a Scribd company logo
1 of 20
© 2015 MapR Technologies 1© 2014 MapR Technologies
© 2015 MapR Technologies 2
Agenda
• Challenges of the Modern Data Center
• Schedulers – Mesos, YARN, and more
• Introducing Apache Myriad
© 2015 MapR Technologies 3
What’s in your data center???
• Your applications
– Tier 1
– Tier 2
– CI/Build
• Primary/secondary storage
• Some databases – relational, NoSQL, and more
• Messaging
• High-priority analytics
• Low-priority best-effort analytics
© 2015 MapR Technologies 4
The utilization problem…
Application Servers Hadoop Servers
Utilizatiion
long lived excess
capacity situations
• Wouldn’t it be nice if you could?
• “Scale up” Hadoop during long periods of low utilization
• “Scale down” Hadoop ahead of anticipated high utilization
© 2015 MapR Technologies 5
What about multi-tenancy?
Tenant
#1
Tenant
#2
Tenant
#3
Cluster #1 Cluster #2 Cluster #3Cluster
Tenant
#1
Tenant
#2
Tenant
#3
Pros:
● No data movement
● Selective data sharing
● Single cluster to manage
● Full use of capacity
Cons:
● Hadoop-only server infrastructure
● Not all Hadoop components fully multi-tenant
● IT defines Hadoop components to offer
Pros:
● Discrete, custom per-tenant Hadoop clusters
● Sharing of infra between Hadoop and other applications
Cons:
● Virtualization impacts performance
● Lots of clusters to manage
● No shared data – additional duplication & movement
● Difficult to do short term “borrowing” of capacity
© 2015 MapR Technologies 6
Schedulers to the rescue! A wish list -
• Applications request resources when they need them
– Without user intervention
• Custom scheduling algorithms –
– Some apps want resources ASAP
– Others want specific resources and are willing to wait
• Multi-tenancy with strong isolation
• Efficient use of resources with preemption & oversubscription
© 2015 MapR Technologies 7
Great. But which one?
YARN
Approach Multi-level Scheduling
Application decides what’s best
Single-level Scheduling
Scheduler decides what’s best
Ideal For Long-lived and short-lived apps Short-lived, task-based jobs
Ecosystem
© 2015 MapR Technologies 8
Mesos Architecture
Mesos
Master
Mesos
Master
Mesos
Master
Myriad
Framework
Marathon
Framework
Mesos Slave Mesos Slave
Myriad
Executor
Mesos
Executor
Mesos
Executor
Docker
Executor
Task ./ruby XYZ java –jar XYZ.jar ./xyz
Tas
k
Zookeeper
Quorum
© 2015 MapR Technologies 9
YARN Architecture
© 2015 MapR Technologies 10
Life with Both
Data
Created
Here
Data
Processed
Here
© 2015 MapR Technologies 11© 2014 MapR Technologies
Apache Myriad (Incubating)
Enables Mesos & YARN to co-exist on
same physical data center infrastructure.
© 2015 MapR Technologies 12
How it works
• Mesos creates virtual clusters
• YARN uses Mesos resources
• YARN can release resources
• Or get more
• Myriad manages conversation
between RM and Mesos
master
– between NM and Mesos too
Mesos
YARN cluster
Web Servers
YARN
cluster
© 2015 MapR Technologies 13
Without using more than it needs
Mesos Master
Myriad
RM
Myriad NM
YARN Task
Myriad NM
YARN Task
Myriad NM
YARN Task
Mesos Slave
Mesos SlaveMesos Slave
Job
© 2015 MapR Technologies 14
What about my storage???
© 2015 MapR Technologies 15
Add some persistent, shared storage
• FS and DB as a service
– Outside the scheduler
• Rapid clusters-on-demand
– Hadoop or other
• Stop cluster, data persists
• Cluster restart doesn’t need
to copy data in
• Share data across clusters
Mesos
Shared Data Services
(FS, DB, …)
© 2015 MapR Technologies 16
Putting it all together
Physical Machines
Distributed Applications
Data Center Operating Services
Data Services
© 2015 MapR Technologies 17
The Future
• Incubator
– Proposal at http://wiki.apache.org/incubator/MyriadProposal
– Initial team from Mesosphere, Paypal, MapR
• Community building
– Diversity is good already
– Starting with very lean team
© 2015 MapR Technologies 18
Myriad Project
• Blog “Project Myriad: No Hadoop is an Island” http://bit.ly/myriad-
mapr-blog
• Proposal to be an incubator project of the Apache Foundation
submitted 12 February 2015 http://bit.ly/myriad-asf-proposal
• Initial code on github: http://bit.ly/github-myriad
• Join us! Twitter for Myriad community @ApacheMyriad
[actual logo coming soon]
© 2015 MapR Technologies 19© 2014 MapR Technologies
Thank You
© 2015 MapR Technologies 20
Myriad Services Architecture
Node ManagerResource Manager
Executor
Mesos
Scheduler
Mesos
Container
Container
App
YARN
Scheduler
(fairshare)
Offers
Launch
Tasks
Launch
Tasks
Task
Status
Launch containers
via HB
Submit
Map<Node,
Capacity>

More Related Content

Viewers also liked

Stackato PaaS Architecture white paper
Stackato PaaS Architecture white paperStackato PaaS Architecture white paper
Stackato PaaS Architecture white paperAngie Hirata
 
8 devstack beyond_hello-world
8 devstack beyond_hello-world8 devstack beyond_hello-world
8 devstack beyond_hello-worldopenstackindia
 
Big Data Day LA 2016/ Hadoop/ Spark/ Kafka track - Iterative Spark Developmen...
Big Data Day LA 2016/ Hadoop/ Spark/ Kafka track - Iterative Spark Developmen...Big Data Day LA 2016/ Hadoop/ Spark/ Kafka track - Iterative Spark Developmen...
Big Data Day LA 2016/ Hadoop/ Spark/ Kafka track - Iterative Spark Developmen...Data Con LA
 
SwiftStack Presents at Under the Radar 2013
SwiftStack Presents at Under the Radar 2013SwiftStack Presents at Under the Radar 2013
SwiftStack Presents at Under the Radar 2013Dealmaker Media
 
Big Data Day LA 2016 Keynote - Andy Feng/ Yahoo
Big Data Day LA 2016 Keynote - Andy Feng/ YahooBig Data Day LA 2016 Keynote - Andy Feng/ Yahoo
Big Data Day LA 2016 Keynote - Andy Feng/ YahooData Con LA
 
Big Data Day LA 2016/ Data Science Track - Backstage to a Data Driven Culture...
Big Data Day LA 2016/ Data Science Track - Backstage to a Data Driven Culture...Big Data Day LA 2016/ Data Science Track - Backstage to a Data Driven Culture...
Big Data Day LA 2016/ Data Science Track - Backstage to a Data Driven Culture...Data Con LA
 
Big Data Day LA 2016/ Big Data Track - Building scalable enterprise data flow...
Big Data Day LA 2016/ Big Data Track - Building scalable enterprise data flow...Big Data Day LA 2016/ Big Data Track - Building scalable enterprise data flow...
Big Data Day LA 2016/ Big Data Track - Building scalable enterprise data flow...Data Con LA
 
Myriad_Product Collaterals
Myriad_Product CollateralsMyriad_Product Collaterals
Myriad_Product CollateralsSuman Mishra
 
Scalable On-Demand Hadoop Clusters with Docker and Mesos
Scalable On-Demand Hadoop Clusters with Docker and MesosScalable On-Demand Hadoop Clusters with Docker and Mesos
Scalable On-Demand Hadoop Clusters with Docker and Mesosnelsonadpresent
 
PaaS Design & Architecture: A Deep Dive into Apache Stratos
PaaS Design & Architecture: A Deep Dive into Apache StratosPaaS Design & Architecture: A Deep Dive into Apache Stratos
PaaS Design & Architecture: A Deep Dive into Apache StratosWSO2
 
From Continous Integration to Continuous Delivery
From Continous Integration to Continuous DeliveryFrom Continous Integration to Continuous Delivery
From Continous Integration to Continuous DeliveryEberhard Wolff
 
Scaling Big Data with Hadoop and Mesos
Scaling Big Data with Hadoop and MesosScaling Big Data with Hadoop and Mesos
Scaling Big Data with Hadoop and MesosDiscover Pinterest
 
Joining the Club: Using Spark to Accelerate Big Data at Dollar Shave Club
Joining the Club: Using Spark to Accelerate Big Data at Dollar Shave ClubJoining the Club: Using Spark to Accelerate Big Data at Dollar Shave Club
Joining the Club: Using Spark to Accelerate Big Data at Dollar Shave ClubData Con LA
 

Viewers also liked (15)

Stackato PaaS Architecture white paper
Stackato PaaS Architecture white paperStackato PaaS Architecture white paper
Stackato PaaS Architecture white paper
 
Apache Mesos
Apache MesosApache Mesos
Apache Mesos
 
8 devstack beyond_hello-world
8 devstack beyond_hello-world8 devstack beyond_hello-world
8 devstack beyond_hello-world
 
Big Data Day LA 2016/ Hadoop/ Spark/ Kafka track - Iterative Spark Developmen...
Big Data Day LA 2016/ Hadoop/ Spark/ Kafka track - Iterative Spark Developmen...Big Data Day LA 2016/ Hadoop/ Spark/ Kafka track - Iterative Spark Developmen...
Big Data Day LA 2016/ Hadoop/ Spark/ Kafka track - Iterative Spark Developmen...
 
SwiftStack Presents at Under the Radar 2013
SwiftStack Presents at Under the Radar 2013SwiftStack Presents at Under the Radar 2013
SwiftStack Presents at Under the Radar 2013
 
Big Data Day LA 2016 Keynote - Andy Feng/ Yahoo
Big Data Day LA 2016 Keynote - Andy Feng/ YahooBig Data Day LA 2016 Keynote - Andy Feng/ Yahoo
Big Data Day LA 2016 Keynote - Andy Feng/ Yahoo
 
Big Data Day LA 2016/ Data Science Track - Backstage to a Data Driven Culture...
Big Data Day LA 2016/ Data Science Track - Backstage to a Data Driven Culture...Big Data Day LA 2016/ Data Science Track - Backstage to a Data Driven Culture...
Big Data Day LA 2016/ Data Science Track - Backstage to a Data Driven Culture...
 
Momentum Myriad
Momentum Myriad Momentum Myriad
Momentum Myriad
 
Big Data Day LA 2016/ Big Data Track - Building scalable enterprise data flow...
Big Data Day LA 2016/ Big Data Track - Building scalable enterprise data flow...Big Data Day LA 2016/ Big Data Track - Building scalable enterprise data flow...
Big Data Day LA 2016/ Big Data Track - Building scalable enterprise data flow...
 
Myriad_Product Collaterals
Myriad_Product CollateralsMyriad_Product Collaterals
Myriad_Product Collaterals
 
Scalable On-Demand Hadoop Clusters with Docker and Mesos
Scalable On-Demand Hadoop Clusters with Docker and MesosScalable On-Demand Hadoop Clusters with Docker and Mesos
Scalable On-Demand Hadoop Clusters with Docker and Mesos
 
PaaS Design & Architecture: A Deep Dive into Apache Stratos
PaaS Design & Architecture: A Deep Dive into Apache StratosPaaS Design & Architecture: A Deep Dive into Apache Stratos
PaaS Design & Architecture: A Deep Dive into Apache Stratos
 
From Continous Integration to Continuous Delivery
From Continous Integration to Continuous DeliveryFrom Continous Integration to Continuous Delivery
From Continous Integration to Continuous Delivery
 
Scaling Big Data with Hadoop and Mesos
Scaling Big Data with Hadoop and MesosScaling Big Data with Hadoop and Mesos
Scaling Big Data with Hadoop and Mesos
 
Joining the Club: Using Spark to Accelerate Big Data at Dollar Shave Club
Joining the Club: Using Spark to Accelerate Big Data at Dollar Shave ClubJoining the Club: Using Spark to Accelerate Big Data at Dollar Shave Club
Joining the Club: Using Spark to Accelerate Big Data at Dollar Shave Club
 

More from Data Con LA

Data Con LA 2022 Keynotes
Data Con LA 2022 KeynotesData Con LA 2022 Keynotes
Data Con LA 2022 KeynotesData Con LA
 
Data Con LA 2022 Keynotes
Data Con LA 2022 KeynotesData Con LA 2022 Keynotes
Data Con LA 2022 KeynotesData Con LA
 
Data Con LA 2022 Keynote
Data Con LA 2022 KeynoteData Con LA 2022 Keynote
Data Con LA 2022 KeynoteData Con LA
 
Data Con LA 2022 - Startup Showcase
Data Con LA 2022 - Startup ShowcaseData Con LA 2022 - Startup Showcase
Data Con LA 2022 - Startup ShowcaseData Con LA
 
Data Con LA 2022 Keynote
Data Con LA 2022 KeynoteData Con LA 2022 Keynote
Data Con LA 2022 KeynoteData Con LA
 
Data Con LA 2022 - Using Google trends data to build product recommendations
Data Con LA 2022 - Using Google trends data to build product recommendationsData Con LA 2022 - Using Google trends data to build product recommendations
Data Con LA 2022 - Using Google trends data to build product recommendationsData Con LA
 
Data Con LA 2022 - AI Ethics
Data Con LA 2022 - AI EthicsData Con LA 2022 - AI Ethics
Data Con LA 2022 - AI EthicsData Con LA
 
Data Con LA 2022 - Improving disaster response with machine learning
Data Con LA 2022 - Improving disaster response with machine learningData Con LA 2022 - Improving disaster response with machine learning
Data Con LA 2022 - Improving disaster response with machine learningData Con LA
 
Data Con LA 2022 - What's new with MongoDB 6.0 and Atlas
Data Con LA 2022 - What's new with MongoDB 6.0 and AtlasData Con LA 2022 - What's new with MongoDB 6.0 and Atlas
Data Con LA 2022 - What's new with MongoDB 6.0 and AtlasData Con LA
 
Data Con LA 2022 - Real world consumer segmentation
Data Con LA 2022 - Real world consumer segmentationData Con LA 2022 - Real world consumer segmentation
Data Con LA 2022 - Real world consumer segmentationData Con LA
 
Data Con LA 2022 - Modernizing Analytics & AI for today's needs: Intuit Turbo...
Data Con LA 2022 - Modernizing Analytics & AI for today's needs: Intuit Turbo...Data Con LA 2022 - Modernizing Analytics & AI for today's needs: Intuit Turbo...
Data Con LA 2022 - Modernizing Analytics & AI for today's needs: Intuit Turbo...Data Con LA
 
Data Con LA 2022 - Moving Data at Scale to AWS
Data Con LA 2022 - Moving Data at Scale to AWSData Con LA 2022 - Moving Data at Scale to AWS
Data Con LA 2022 - Moving Data at Scale to AWSData Con LA
 
Data Con LA 2022 - Collaborative Data Exploration using Conversational AI
Data Con LA 2022 - Collaborative Data Exploration using Conversational AIData Con LA 2022 - Collaborative Data Exploration using Conversational AI
Data Con LA 2022 - Collaborative Data Exploration using Conversational AIData Con LA
 
Data Con LA 2022 - Why Database Modernization Makes Your Data Decisions More ...
Data Con LA 2022 - Why Database Modernization Makes Your Data Decisions More ...Data Con LA 2022 - Why Database Modernization Makes Your Data Decisions More ...
Data Con LA 2022 - Why Database Modernization Makes Your Data Decisions More ...Data Con LA
 
Data Con LA 2022 - Intro to Data Science
Data Con LA 2022 - Intro to Data ScienceData Con LA 2022 - Intro to Data Science
Data Con LA 2022 - Intro to Data ScienceData Con LA
 
Data Con LA 2022 - How are NFTs and DeFi Changing Entertainment
Data Con LA 2022 - How are NFTs and DeFi Changing EntertainmentData Con LA 2022 - How are NFTs and DeFi Changing Entertainment
Data Con LA 2022 - How are NFTs and DeFi Changing EntertainmentData Con LA
 
Data Con LA 2022 - Why Data Quality vigilance requires an End-to-End, Automat...
Data Con LA 2022 - Why Data Quality vigilance requires an End-to-End, Automat...Data Con LA 2022 - Why Data Quality vigilance requires an End-to-End, Automat...
Data Con LA 2022 - Why Data Quality vigilance requires an End-to-End, Automat...Data Con LA
 
Data Con LA 2022-Perfect Viral Ad prediction of Superbowl 2022 using Tease, T...
Data Con LA 2022-Perfect Viral Ad prediction of Superbowl 2022 using Tease, T...Data Con LA 2022-Perfect Viral Ad prediction of Superbowl 2022 using Tease, T...
Data Con LA 2022-Perfect Viral Ad prediction of Superbowl 2022 using Tease, T...Data Con LA
 
Data Con LA 2022- Embedding medical journeys with machine learning to improve...
Data Con LA 2022- Embedding medical journeys with machine learning to improve...Data Con LA 2022- Embedding medical journeys with machine learning to improve...
Data Con LA 2022- Embedding medical journeys with machine learning to improve...Data Con LA
 
Data Con LA 2022 - Data Streaming with Kafka
Data Con LA 2022 - Data Streaming with KafkaData Con LA 2022 - Data Streaming with Kafka
Data Con LA 2022 - Data Streaming with KafkaData Con LA
 

More from Data Con LA (20)

Data Con LA 2022 Keynotes
Data Con LA 2022 KeynotesData Con LA 2022 Keynotes
Data Con LA 2022 Keynotes
 
Data Con LA 2022 Keynotes
Data Con LA 2022 KeynotesData Con LA 2022 Keynotes
Data Con LA 2022 Keynotes
 
Data Con LA 2022 Keynote
Data Con LA 2022 KeynoteData Con LA 2022 Keynote
Data Con LA 2022 Keynote
 
Data Con LA 2022 - Startup Showcase
Data Con LA 2022 - Startup ShowcaseData Con LA 2022 - Startup Showcase
Data Con LA 2022 - Startup Showcase
 
Data Con LA 2022 Keynote
Data Con LA 2022 KeynoteData Con LA 2022 Keynote
Data Con LA 2022 Keynote
 
Data Con LA 2022 - Using Google trends data to build product recommendations
Data Con LA 2022 - Using Google trends data to build product recommendationsData Con LA 2022 - Using Google trends data to build product recommendations
Data Con LA 2022 - Using Google trends data to build product recommendations
 
Data Con LA 2022 - AI Ethics
Data Con LA 2022 - AI EthicsData Con LA 2022 - AI Ethics
Data Con LA 2022 - AI Ethics
 
Data Con LA 2022 - Improving disaster response with machine learning
Data Con LA 2022 - Improving disaster response with machine learningData Con LA 2022 - Improving disaster response with machine learning
Data Con LA 2022 - Improving disaster response with machine learning
 
Data Con LA 2022 - What's new with MongoDB 6.0 and Atlas
Data Con LA 2022 - What's new with MongoDB 6.0 and AtlasData Con LA 2022 - What's new with MongoDB 6.0 and Atlas
Data Con LA 2022 - What's new with MongoDB 6.0 and Atlas
 
Data Con LA 2022 - Real world consumer segmentation
Data Con LA 2022 - Real world consumer segmentationData Con LA 2022 - Real world consumer segmentation
Data Con LA 2022 - Real world consumer segmentation
 
Data Con LA 2022 - Modernizing Analytics & AI for today's needs: Intuit Turbo...
Data Con LA 2022 - Modernizing Analytics & AI for today's needs: Intuit Turbo...Data Con LA 2022 - Modernizing Analytics & AI for today's needs: Intuit Turbo...
Data Con LA 2022 - Modernizing Analytics & AI for today's needs: Intuit Turbo...
 
Data Con LA 2022 - Moving Data at Scale to AWS
Data Con LA 2022 - Moving Data at Scale to AWSData Con LA 2022 - Moving Data at Scale to AWS
Data Con LA 2022 - Moving Data at Scale to AWS
 
Data Con LA 2022 - Collaborative Data Exploration using Conversational AI
Data Con LA 2022 - Collaborative Data Exploration using Conversational AIData Con LA 2022 - Collaborative Data Exploration using Conversational AI
Data Con LA 2022 - Collaborative Data Exploration using Conversational AI
 
Data Con LA 2022 - Why Database Modernization Makes Your Data Decisions More ...
Data Con LA 2022 - Why Database Modernization Makes Your Data Decisions More ...Data Con LA 2022 - Why Database Modernization Makes Your Data Decisions More ...
Data Con LA 2022 - Why Database Modernization Makes Your Data Decisions More ...
 
Data Con LA 2022 - Intro to Data Science
Data Con LA 2022 - Intro to Data ScienceData Con LA 2022 - Intro to Data Science
Data Con LA 2022 - Intro to Data Science
 
Data Con LA 2022 - How are NFTs and DeFi Changing Entertainment
Data Con LA 2022 - How are NFTs and DeFi Changing EntertainmentData Con LA 2022 - How are NFTs and DeFi Changing Entertainment
Data Con LA 2022 - How are NFTs and DeFi Changing Entertainment
 
Data Con LA 2022 - Why Data Quality vigilance requires an End-to-End, Automat...
Data Con LA 2022 - Why Data Quality vigilance requires an End-to-End, Automat...Data Con LA 2022 - Why Data Quality vigilance requires an End-to-End, Automat...
Data Con LA 2022 - Why Data Quality vigilance requires an End-to-End, Automat...
 
Data Con LA 2022-Perfect Viral Ad prediction of Superbowl 2022 using Tease, T...
Data Con LA 2022-Perfect Viral Ad prediction of Superbowl 2022 using Tease, T...Data Con LA 2022-Perfect Viral Ad prediction of Superbowl 2022 using Tease, T...
Data Con LA 2022-Perfect Viral Ad prediction of Superbowl 2022 using Tease, T...
 
Data Con LA 2022- Embedding medical journeys with machine learning to improve...
Data Con LA 2022- Embedding medical journeys with machine learning to improve...Data Con LA 2022- Embedding medical journeys with machine learning to improve...
Data Con LA 2022- Embedding medical journeys with machine learning to improve...
 
Data Con LA 2022 - Data Streaming with Kafka
Data Con LA 2022 - Data Streaming with KafkaData Con LA 2022 - Data Streaming with Kafka
Data Con LA 2022 - Data Streaming with Kafka
 

Recently uploaded

Designing for Hardware Accessibility at Comcast
Designing for Hardware Accessibility at ComcastDesigning for Hardware Accessibility at Comcast
Designing for Hardware Accessibility at ComcastUXDXConf
 
Long journey of Ruby Standard library at RubyKaigi 2024
Long journey of Ruby Standard library at RubyKaigi 2024Long journey of Ruby Standard library at RubyKaigi 2024
Long journey of Ruby Standard library at RubyKaigi 2024Hiroshi SHIBATA
 
ECS 2024 Teams Premium - Pretty Secure
ECS 2024   Teams Premium - Pretty SecureECS 2024   Teams Premium - Pretty Secure
ECS 2024 Teams Premium - Pretty SecureFemke de Vroome
 
How we scaled to 80K users by doing nothing!.pdf
How we scaled to 80K users by doing nothing!.pdfHow we scaled to 80K users by doing nothing!.pdf
How we scaled to 80K users by doing nothing!.pdfSrushith Repakula
 
Behind the Scenes From the Manager's Chair: Decoding the Secrets of Successfu...
Behind the Scenes From the Manager's Chair: Decoding the Secrets of Successfu...Behind the Scenes From the Manager's Chair: Decoding the Secrets of Successfu...
Behind the Scenes From the Manager's Chair: Decoding the Secrets of Successfu...CzechDreamin
 
State of the Smart Building Startup Landscape 2024!
State of the Smart Building Startup Landscape 2024!State of the Smart Building Startup Landscape 2024!
State of the Smart Building Startup Landscape 2024!Memoori
 
How Red Hat Uses FDO in Device Lifecycle _ Costin and Vitaliy at Red Hat.pdf
How Red Hat Uses FDO in Device Lifecycle _ Costin and Vitaliy at Red Hat.pdfHow Red Hat Uses FDO in Device Lifecycle _ Costin and Vitaliy at Red Hat.pdf
How Red Hat Uses FDO in Device Lifecycle _ Costin and Vitaliy at Red Hat.pdfFIDO Alliance
 
Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...
Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...
Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...FIDO Alliance
 
Working together SRE & Platform Engineering
Working together SRE & Platform EngineeringWorking together SRE & Platform Engineering
Working together SRE & Platform EngineeringMarcus Vechiato
 
WebAssembly is Key to Better LLM Performance
WebAssembly is Key to Better LLM PerformanceWebAssembly is Key to Better LLM Performance
WebAssembly is Key to Better LLM PerformanceSamy Fodil
 
Breaking Down the Flutterwave Scandal What You Need to Know.pdf
Breaking Down the Flutterwave Scandal What You Need to Know.pdfBreaking Down the Flutterwave Scandal What You Need to Know.pdf
Breaking Down the Flutterwave Scandal What You Need to Know.pdfUK Journal
 
Extensible Python: Robustness through Addition - PyCon 2024
Extensible Python: Robustness through Addition - PyCon 2024Extensible Python: Robustness through Addition - PyCon 2024
Extensible Python: Robustness through Addition - PyCon 2024Patrick Viafore
 
Google I/O Extended 2024 Warsaw
Google I/O Extended 2024 WarsawGoogle I/O Extended 2024 Warsaw
Google I/O Extended 2024 WarsawGDSC PJATK
 
Using IESVE for Room Loads Analysis - UK & Ireland
Using IESVE for Room Loads Analysis - UK & IrelandUsing IESVE for Room Loads Analysis - UK & Ireland
Using IESVE for Room Loads Analysis - UK & IrelandIES VE
 
TopCryptoSupers 12thReport OrionX May2024
TopCryptoSupers 12thReport OrionX May2024TopCryptoSupers 12thReport OrionX May2024
TopCryptoSupers 12thReport OrionX May2024Stephen Perrenod
 
Structuring Teams and Portfolios for Success
Structuring Teams and Portfolios for SuccessStructuring Teams and Portfolios for Success
Structuring Teams and Portfolios for SuccessUXDXConf
 
PLAI - Acceleration Program for Generative A.I. Startups
PLAI - Acceleration Program for Generative A.I. StartupsPLAI - Acceleration Program for Generative A.I. Startups
PLAI - Acceleration Program for Generative A.I. StartupsStefano
 
BT & Neo4j _ How Knowledge Graphs help BT deliver Digital Transformation.pptx
BT & Neo4j _ How Knowledge Graphs help BT deliver Digital Transformation.pptxBT & Neo4j _ How Knowledge Graphs help BT deliver Digital Transformation.pptx
BT & Neo4j _ How Knowledge Graphs help BT deliver Digital Transformation.pptxNeo4j
 
The Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdf
The Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdfThe Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdf
The Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdfFIDO Alliance
 
Your enemies use GenAI too - staying ahead of fraud with Neo4j
Your enemies use GenAI too - staying ahead of fraud with Neo4jYour enemies use GenAI too - staying ahead of fraud with Neo4j
Your enemies use GenAI too - staying ahead of fraud with Neo4jNeo4j
 

Recently uploaded (20)

Designing for Hardware Accessibility at Comcast
Designing for Hardware Accessibility at ComcastDesigning for Hardware Accessibility at Comcast
Designing for Hardware Accessibility at Comcast
 
Long journey of Ruby Standard library at RubyKaigi 2024
Long journey of Ruby Standard library at RubyKaigi 2024Long journey of Ruby Standard library at RubyKaigi 2024
Long journey of Ruby Standard library at RubyKaigi 2024
 
ECS 2024 Teams Premium - Pretty Secure
ECS 2024   Teams Premium - Pretty SecureECS 2024   Teams Premium - Pretty Secure
ECS 2024 Teams Premium - Pretty Secure
 
How we scaled to 80K users by doing nothing!.pdf
How we scaled to 80K users by doing nothing!.pdfHow we scaled to 80K users by doing nothing!.pdf
How we scaled to 80K users by doing nothing!.pdf
 
Behind the Scenes From the Manager's Chair: Decoding the Secrets of Successfu...
Behind the Scenes From the Manager's Chair: Decoding the Secrets of Successfu...Behind the Scenes From the Manager's Chair: Decoding the Secrets of Successfu...
Behind the Scenes From the Manager's Chair: Decoding the Secrets of Successfu...
 
State of the Smart Building Startup Landscape 2024!
State of the Smart Building Startup Landscape 2024!State of the Smart Building Startup Landscape 2024!
State of the Smart Building Startup Landscape 2024!
 
How Red Hat Uses FDO in Device Lifecycle _ Costin and Vitaliy at Red Hat.pdf
How Red Hat Uses FDO in Device Lifecycle _ Costin and Vitaliy at Red Hat.pdfHow Red Hat Uses FDO in Device Lifecycle _ Costin and Vitaliy at Red Hat.pdf
How Red Hat Uses FDO in Device Lifecycle _ Costin and Vitaliy at Red Hat.pdf
 
Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...
Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...
Secure Zero Touch enabled Edge compute with Dell NativeEdge via FDO _ Brad at...
 
Working together SRE & Platform Engineering
Working together SRE & Platform EngineeringWorking together SRE & Platform Engineering
Working together SRE & Platform Engineering
 
WebAssembly is Key to Better LLM Performance
WebAssembly is Key to Better LLM PerformanceWebAssembly is Key to Better LLM Performance
WebAssembly is Key to Better LLM Performance
 
Breaking Down the Flutterwave Scandal What You Need to Know.pdf
Breaking Down the Flutterwave Scandal What You Need to Know.pdfBreaking Down the Flutterwave Scandal What You Need to Know.pdf
Breaking Down the Flutterwave Scandal What You Need to Know.pdf
 
Extensible Python: Robustness through Addition - PyCon 2024
Extensible Python: Robustness through Addition - PyCon 2024Extensible Python: Robustness through Addition - PyCon 2024
Extensible Python: Robustness through Addition - PyCon 2024
 
Google I/O Extended 2024 Warsaw
Google I/O Extended 2024 WarsawGoogle I/O Extended 2024 Warsaw
Google I/O Extended 2024 Warsaw
 
Using IESVE for Room Loads Analysis - UK & Ireland
Using IESVE for Room Loads Analysis - UK & IrelandUsing IESVE for Room Loads Analysis - UK & Ireland
Using IESVE for Room Loads Analysis - UK & Ireland
 
TopCryptoSupers 12thReport OrionX May2024
TopCryptoSupers 12thReport OrionX May2024TopCryptoSupers 12thReport OrionX May2024
TopCryptoSupers 12thReport OrionX May2024
 
Structuring Teams and Portfolios for Success
Structuring Teams and Portfolios for SuccessStructuring Teams and Portfolios for Success
Structuring Teams and Portfolios for Success
 
PLAI - Acceleration Program for Generative A.I. Startups
PLAI - Acceleration Program for Generative A.I. StartupsPLAI - Acceleration Program for Generative A.I. Startups
PLAI - Acceleration Program for Generative A.I. Startups
 
BT & Neo4j _ How Knowledge Graphs help BT deliver Digital Transformation.pptx
BT & Neo4j _ How Knowledge Graphs help BT deliver Digital Transformation.pptxBT & Neo4j _ How Knowledge Graphs help BT deliver Digital Transformation.pptx
BT & Neo4j _ How Knowledge Graphs help BT deliver Digital Transformation.pptx
 
The Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdf
The Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdfThe Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdf
The Value of Certifying Products for FDO _ Paul at FIDO Alliance.pdf
 
Your enemies use GenAI too - staying ahead of fraud with Neo4j
Your enemies use GenAI too - staying ahead of fraud with Neo4jYour enemies use GenAI too - staying ahead of fraud with Neo4j
Your enemies use GenAI too - staying ahead of fraud with Neo4j
 

Big Data Day LA 2015 - Introducing Myriad, a Mesos framework for dynamically scaling Hadoop workloads by Will Ochandarena of MapR

  • 1. © 2015 MapR Technologies 1© 2014 MapR Technologies
  • 2. © 2015 MapR Technologies 2 Agenda • Challenges of the Modern Data Center • Schedulers – Mesos, YARN, and more • Introducing Apache Myriad
  • 3. © 2015 MapR Technologies 3 What’s in your data center??? • Your applications – Tier 1 – Tier 2 – CI/Build • Primary/secondary storage • Some databases – relational, NoSQL, and more • Messaging • High-priority analytics • Low-priority best-effort analytics
  • 4. © 2015 MapR Technologies 4 The utilization problem… Application Servers Hadoop Servers Utilizatiion long lived excess capacity situations • Wouldn’t it be nice if you could? • “Scale up” Hadoop during long periods of low utilization • “Scale down” Hadoop ahead of anticipated high utilization
  • 5. © 2015 MapR Technologies 5 What about multi-tenancy? Tenant #1 Tenant #2 Tenant #3 Cluster #1 Cluster #2 Cluster #3Cluster Tenant #1 Tenant #2 Tenant #3 Pros: ● No data movement ● Selective data sharing ● Single cluster to manage ● Full use of capacity Cons: ● Hadoop-only server infrastructure ● Not all Hadoop components fully multi-tenant ● IT defines Hadoop components to offer Pros: ● Discrete, custom per-tenant Hadoop clusters ● Sharing of infra between Hadoop and other applications Cons: ● Virtualization impacts performance ● Lots of clusters to manage ● No shared data – additional duplication & movement ● Difficult to do short term “borrowing” of capacity
  • 6. © 2015 MapR Technologies 6 Schedulers to the rescue! A wish list - • Applications request resources when they need them – Without user intervention • Custom scheduling algorithms – – Some apps want resources ASAP – Others want specific resources and are willing to wait • Multi-tenancy with strong isolation • Efficient use of resources with preemption & oversubscription
  • 7. © 2015 MapR Technologies 7 Great. But which one? YARN Approach Multi-level Scheduling Application decides what’s best Single-level Scheduling Scheduler decides what’s best Ideal For Long-lived and short-lived apps Short-lived, task-based jobs Ecosystem
  • 8. © 2015 MapR Technologies 8 Mesos Architecture Mesos Master Mesos Master Mesos Master Myriad Framework Marathon Framework Mesos Slave Mesos Slave Myriad Executor Mesos Executor Mesos Executor Docker Executor Task ./ruby XYZ java –jar XYZ.jar ./xyz Tas k Zookeeper Quorum
  • 9. © 2015 MapR Technologies 9 YARN Architecture
  • 10. © 2015 MapR Technologies 10 Life with Both Data Created Here Data Processed Here
  • 11. © 2015 MapR Technologies 11© 2014 MapR Technologies Apache Myriad (Incubating) Enables Mesos & YARN to co-exist on same physical data center infrastructure.
  • 12. © 2015 MapR Technologies 12 How it works • Mesos creates virtual clusters • YARN uses Mesos resources • YARN can release resources • Or get more • Myriad manages conversation between RM and Mesos master – between NM and Mesos too Mesos YARN cluster Web Servers YARN cluster
  • 13. © 2015 MapR Technologies 13 Without using more than it needs Mesos Master Myriad RM Myriad NM YARN Task Myriad NM YARN Task Myriad NM YARN Task Mesos Slave Mesos SlaveMesos Slave Job
  • 14. © 2015 MapR Technologies 14 What about my storage???
  • 15. © 2015 MapR Technologies 15 Add some persistent, shared storage • FS and DB as a service – Outside the scheduler • Rapid clusters-on-demand – Hadoop or other • Stop cluster, data persists • Cluster restart doesn’t need to copy data in • Share data across clusters Mesos Shared Data Services (FS, DB, …)
  • 16. © 2015 MapR Technologies 16 Putting it all together Physical Machines Distributed Applications Data Center Operating Services Data Services
  • 17. © 2015 MapR Technologies 17 The Future • Incubator – Proposal at http://wiki.apache.org/incubator/MyriadProposal – Initial team from Mesosphere, Paypal, MapR • Community building – Diversity is good already – Starting with very lean team
  • 18. © 2015 MapR Technologies 18 Myriad Project • Blog “Project Myriad: No Hadoop is an Island” http://bit.ly/myriad- mapr-blog • Proposal to be an incubator project of the Apache Foundation submitted 12 February 2015 http://bit.ly/myriad-asf-proposal • Initial code on github: http://bit.ly/github-myriad • Join us! Twitter for Myriad community @ApacheMyriad [actual logo coming soon]
  • 19. © 2015 MapR Technologies 19© 2014 MapR Technologies Thank You
  • 20. © 2015 MapR Technologies 20 Myriad Services Architecture Node ManagerResource Manager Executor Mesos Scheduler Mesos Container Container App YARN Scheduler (fairshare) Offers Launch Tasks Launch Tasks Task Status Launch containers via HB Submit Map<Node, Capacity>