leveraging nosql in the cloud

20
1 Leveraging NoSQL in the Cloud Uri Budnik (RightScale) Frank Weigel (Couchbase)

Upload: couchbase

Post on 20-Aug-2015

533 views

Category:

Technology


3 download

TRANSCRIPT

Page 1: Leveraging NoSQL in the Cloud

1  

Leveraging  NoSQL  in  the  Cloud    

Uri  Budnik  (RightScale)  Frank  Weigel  (Couchbase)  

 

Page 2: Leveraging NoSQL in the Cloud

# 2 # 2

RightScale Pioneered Cloud Management

•  World’s cloud management platform

•  A shop for managing the entire stack

•  Managing cloud deployments globally for

• 

•  More than launched

•  Powering the production cloud deployments

•  fully supported cloud providers and counting

Page 3: Leveraging NoSQL in the Cloud

# 3 # 3

•  Four horsemen •  Cut costs •  Improve agility •  Increase scale & reliability •  Leverage best practices

•  Highly effective layers of abstraction •  SaaS – cloaked below the browser •  PaaS – cloaked below the code •  IaaS – cloaked below the OS

•  Each layer achieves scalability & reliability through uniformity •  True IaaS clouds are API-driven

What’s Driving Cloud?

Page 4: Leveraging NoSQL in the Cloud

# 4 # 4

So what’s the problem?

•  Paying attention to design •  Using cloud != designing for cloud •  Accidental tourist in cloud

•  Sign up, launch servers, sit back

•  Good cloud design •  Secure •  Reliable •  No lock-in •  Fault-tolerance •  Scalability

Page 5: Leveraging NoSQL in the Cloud

# 5 # 5

The DIY Trap •  Most clouds are a set of APIs

and/or a simple UI to launch servers – is that what you need?

•  These are basic building blocks, not a management system •  How should you spend your time?

•  Managing multiple users with different levels of access •  Configuration management and app lifecycle management •  Track usage and costs across applications and business units •  All the general purpose things you need to do like: Monitoring, Alarms,

Auto-scaling, etc

What do you do about IT or development personnel unilaterally opening a cloud account and deploying an environment?

Page 6: Leveraging NoSQL in the Cloud

# 6 # 6

Why you should care about multi-cloud •  Multi-cloud scenarios

•  Within / between public clouds •  Public / private hybrid

•  Lock-in dangers •  Data has mass •  Vendor contracts, features & APIs

•  High availability and disaster recovery •  What you should do

•  Preserve multiple choices for any service •  Keep contracts to length you can predict •  Avoid unique services & APIs

•  Anticipate moving – set up small footprints elsewhere •  Pilot a private cloud •  Design for portability: architecture choices matter

Page 7: Leveraging NoSQL in the Cloud

# 7 # 7

One Cloud – Many Resource Pools

Rackspace

Amazon Web Services

SoftLayer

Logicworks

Datapipe

Singapore

Hong Kong

Tokyo

Dallas

DC Area

NYC Area

SF Area

Seattle

Chicago

Dublin

London Amsterdam

IDCF / Yahoo! Japan

Fukuoka

Korea Telecom

Seoul

Beijing

UnGeo

Tata

Hyderabad

Private Clouds Public & Managed Clouds

Oregon

São Paulo

KINX

Page 8: Leveraging NoSQL in the Cloud

# 8 # 8

Page 9: Leveraging NoSQL in the Cloud

9  

ARCHITECTING  APPLICATIONS  FOR  THE  CLOUD  

Page 10: Leveraging NoSQL in the Cloud

10  

Cloud  Environment  ProperCes  

•  ParCal  network  outages  •  Complete  node  failures  happen  •  Limited  node  resources  compared  to  dedicated  datacenter    –  IO  limited  –  CPU  and  RAM  

•  Node  resources  must  be  used  both  efficiently  and  flexibly  

•  SoluCons  must  be  architected  to  fit  the  Cloud  

Page 11: Leveraging NoSQL in the Cloud

11  

Typical  Cloud  Architecture  Stack  

Page 12: Leveraging NoSQL in the Cloud

12  

Scaling  the  ApplicaCon  Layer  is  Easy  

•  ApplicaCon  Servers  hold  relevant  state  in  Data  Layer  •  Scale  by  adding  more  applicaCon  servers  •  Load  balancing  Cer  selects  applicaCon  server    

Page 13: Leveraging NoSQL in the Cloud

13  

Scaling  the  ApplicaCon  Layer  is  Efficient  

Page 14: Leveraging NoSQL in the Cloud

14  

RDBMS  designed  for  different  use-­‐cases  

Page 15: Leveraging NoSQL in the Cloud

15  

Data  Layer  Scaling  hard  with  RDBMS  

•  Limited  scale  up  in  the  cloud  •  RDBMS  do  not  scale  out  •  Cost  explosion  &  Latency  “explosion”  

Page 16: Leveraging NoSQL in the Cloud

16  

NoSQL  Databases  

•  New generation of databases, designed to address needs of modern web & cloud focused solutions

•  No fixed schema –  No more costly schema updates

•  Automatic key sharding

•  Low latency optimized

•  No multi-operation transaction support

Page 17: Leveraging NoSQL in the Cloud

17  

Couchbase  Server  Data  Layer  scales  like  ApplicaCon  Layer  Data  layer  now  scales  with  linear  cost  and  constant  performance.  

•  Scale by adding more database servers •  Sharding to correct server based on key

Page 18: Leveraging NoSQL in the Cloud

18  

Scaling  with  Couchbase  is  easy!  

•  Horizontal scalability of Data Tier •  Linear cost •  Flat response times

Page 19: Leveraging NoSQL in the Cloud

19  

Summary  

•  Done  right,  moving  to  the  cloud  offers  significant  advantages  –  Cost  reducCon,  beaer  agility,  increased  scalability  &  reliability  

•  Managing  and  monitoring  deployments  in  the  cloud  greatly  benefit  from  using  a  cloud  management  plaborm  on  top  of  the  IaaS  APIs.  

•  Make  sure  your  data  storage  layer  is  designed  for  the  cloud  –  Horizontally  scalable,  fault-­‐tolerant  &  high  performance  

Page 20: Leveraging NoSQL in the Cloud

20  

Q&A    

@FRANKWEIGEL                                        @URIBUDNIK  [email protected]          [email protected]  

     

Explore  a  ready-­‐to-­‐launch  example  applicaCon  stack  hap://www.couchbase.com/fast-­‐track-­‐your-­‐app