Page 1
Intelligent People. Uncommon Ideas.
11
Building a Scalable Architecture for Web Apps - Part I
(Lessons Learned @ Directi)
By Bhavin Turakhia CEO, Directi
(http://www.directi.com | http://wiki.directi.com | http://careers.directi.com)
Licensed under Creative Commons Attribution Sharealike Noncommercial
Page 2
22Creative Commons Sharealike Attributions Noncommercial
Agenda
• Why is Scalability important
• Introduction to the Variables and Factors
• Building our own Scalable Architecture (in incremental steps) Vertical Scaling Vertical Partitioning Horizontal Scaling Horizontal Partitioning … etc
• Platform Selection Considerations
• Tips
Page 3
33Creative Commons Sharealike Attributions Noncommercial
Why is Scalability Important in a Web 2.0 world
• Viral marketing can result in instant successes
• RSS / Ajax / SOA pull based / polling type XML protocols - Meta-data > data Number of Requests exponentially grows with user base
• RoR / Grails – Dynamic language landscape gaining popularity
• In the end you want to build a Web 2.0 app that can serve millions of users with ZERO downtime
Page 4
44Creative Commons Sharealike Attributions Noncommercial
The Variables
• Scalability - Number of users / sessions / transactions / operations the entire system can perform
• Performance – Optimal utilization of resources
• Responsiveness – Time taken per operation
• Availability - Probability of the application or a portion of the application being available at any given point in time
• Downtime Impact - The impact of a downtime of a server/service/resource - number of users, type of impact etc
• Cost
• Maintenance Effort
High: scalability, availability, performance & responsivenessLow: downtime impact, cost & maintenance effort
Page 5
55Creative Commons Sharealike Attributions Noncommercial
The Factors
• Platform selection
• Hardware
• Application Design
• Database/Datastore Structure and Architecture
• Deployment Architecture
• Storage Architecture
• Abuse prevention
• Monitoring mechanisms
• … and more
Page 6
66Creative Commons Sharealike Attributions Noncommercial
Lets Start …
• We will now build an example architecture for an example app using the following iterative incremental steps – Inspect current Architecture Identify Scalability Bottlenecks Identify SPOFs and Availability Issues Identify Downtime Impact Risk Zones Apply one of -
• Vertical Scaling
• Vertical Partitioning
• Horizontal Scaling
• Horizontal Partitioning
Repeat process
Page 7
77Creative Commons Sharealike Attributions Noncommercial
Step 1 – Lets Start …
Appserver & DBServer
Page 8
88Creative Commons Sharealike Attributions Noncommercial
Step 2 – Vertical Scaling
Appserver, DBServer
CPU
CPURAM RAM
Page 9
99Creative Commons Sharealike Attributions Noncommercial
Step 2 - Vertical Scaling
• Introduction Increasing the hardware resources
without changing the number of nodes Referred to as “Scaling up” the Server
• Advantages Simple to implement
• Disadvantages Finite limit Hardware does not scale linearly
(diminishing returns for each incremental unit)
Requires downtime Increases Downtime Impact Incremental costs increase
exponentially
Appserver, DBServer
CPU
CPURAM RAM
CPU
CPU
RAM RAM
Page 10
1100Creative Commons Sharealike Attributions Noncommercial
Step 3 – Vertical Partitioning (Services)
AppServer
DBServer
• Introduction Deploying each service on a separate node
• Positives Increases per application Availability Task-based specialization, optimization and
tuning possible Reduces context switching Simple to implement for out of band
processes No changes to App required Flexibility increases
• Negatives Sub-optimal resource utilization May not increase overall availability Finite Scalability
Page 11
1111Creative Commons Sharealike Attributions Noncommercial
Understanding Vertical Partitioning
• The term Vertical Partitioning denotes – Increase in the number of nodes by distributing the
tasks/functions Each node (or cluster) performs separate Tasks Each node (or cluster) is different from the other
• Vertical Partitioning can be performed at various layers (App / Server / Data / Hardware etc)
Page 12
1122Creative Commons Sharealike Attributions Noncommercial
Step 4 – Horizontal Scaling (App Server)
AppServer AppServer AppServer
Load Balancer
DBServer
• Introduction Increasing the number of nodes of
the App Server through Load Balancing
Referred to as “Scaling out” the App Server
Page 13
1133Creative Commons Sharealike Attributions Noncommercial
Understanding Horizontal Scaling
• The term Horizontal Scaling denotes – Increase in the number of nodes by replicating the nodes Each node performs the same Tasks Each node is identical Typically the collection of nodes maybe known as a cluster
(though the term cluster is often misused) Also referred to as “Scaling Out”
• Horizontal Scaling can be performed for any particular type of node (AppServer / DBServer etc)
Page 14
1144Creative Commons Sharealike Attributions Noncommercial
Load Balancer – Hardware vs Software
• Hardware Load balancers are faster• Software Load balancers are more customizable• With HTTP Servers load balancing is typically combined
with http accelerators
Page 15
1155Creative Commons Sharealike Attributions Noncommercial
Load Balancer – Session Management
• Sticky Sessions Requests for a given user are
sent to a fixed App Server Observations
• Asymmetrical load distribution (especially during downtimes)
• Downtime Impact – Loss of session data
AppServer AppServer AppServer
Load Balancer
Sticky Sessions
User 1 User 2
Page 16
1166Creative Commons Sharealike Attributions Noncommercial
Load Balancer – Session Management
• Central Session Store Introduces SPOF An additional variable Session reads and writes
generate Disk + Network I/O Also known as a Shared
Session Store Cluster
AppServer AppServer AppServer
Load Balancer
Session Store
Central Session Storage
Page 17
1177Creative Commons Sharealike Attributions Noncommercial
Load Balancer – Session Management
• Clustered Session Management Easier to setup No SPOF Session reads are instantaneous Session writes generate Network
I/O Network I/O increases
exponentially with increase in number of nodes
In very rare circumstances a request may get stale session data
• User request reaches subsequent node faster than intra-node message
• Intra-node communication fails AKA Shared-nothing Cluster
AppServer AppServer AppServer
Load Balancer
Clustered Session Management
Page 18
1188Creative Commons Sharealike Attributions Noncommercial
Load Balancer – Session Management
• Sticky Sessions with Central Session Store Downtime does not cause loss
of data Session reads need not
generate network I/O
• Sticky Sessions with Clustered Session Management No specific advantages
Sticky Sessions
AppServer AppServer AppServer
Load Balancer
User 1 User 2
Page 19
1199Creative Commons Sharealike Attributions Noncommercial
Load Balancer – Session Management
• Recommendation Use Clustered Session Management if you have –
• Smaller Number of App Servers
• Fewer Session writes Use a Central Session Store elsewhere Use sticky sessions only if you have to
Page 20
2200Creative Commons Sharealike Attributions Noncommercial
Load Balancer – Removing SPOF
• In a Load Balanced App Server Cluster the LB is an SPOF
• Setup LB in Active-Active or Active-Passive mode Note: Active-Active nevertheless
assumes that each LB is independently able to take up the load of the other
If one wants ZERO downtime, then Active-Active becomes truly cost beneficial only if multiple LBs (more than 3 to 4) are daisy chained as Active-Active forming an LB Cluster
AppServer AppServer AppServer
Load Balancer
Active-Passive LB
Load Balancer
AppServer AppServer AppServer
Load Balancer
Active-Active LB
Load Balancer
Users
Users
Page 21
2211Creative Commons Sharealike Attributions Noncommercial
Step 4 – Horizontal Scaling (App Server)
DBServer
• Our deployment at the end of Step 4
• Positives Increases Availability and
Scalability No changes to App required Easy setup
• Negatives Finite Scalability
Load Balanced App Servers
Page 22
2222Creative Commons Sharealike Attributions Noncommercial
Step 5 – Vertical Partitioning (Hardware)
DBServer
• Introduction Partitioning out the Storage
function using a SAN
• SAN config options Refer to “Demystifying Storage” at
http://wiki.directi.com -> Dev University -> Presentations
• Positives Allows “Scaling Up” the DB Server Boosts Performance of DB Server
• Negatives Increases Cost
Load Balanced App Servers
SAN
Page 23
2233Creative Commons Sharealike Attributions Noncommercial
Step 6 – Horizontal Scaling (DB)
DBServer
• Introduction Increasing the number of DB nodes Referred to as “Scaling out” the DB
Server
• Options Shared nothing Cluster Real Application Cluster (or Shared
Storage Cluster)DBServer DBServer
Load Balanced App Servers
SAN
Page 24
2244Creative Commons Sharealike Attributions Noncommercial
Shared Nothing Cluster
• Each DB Server node has its own complete copy of the database
• Nothing is shared between the DB Server Nodes
• This is achieved through DB Replication at DB / Driver / App level or through a proxy
• Supported by most RDBMs natively or through 3rd party software
DBServer
Database
DBServer
Database
DBServer
Database
Note: Actual DB files maybe stored on a central SAN
Page 25
2255Creative Commons Sharealike Attributions Noncommercial
Replication Considerations
• Master-Slave Writes are sent to a single master which replicates the data to
multiple slave nodes Replication maybe cascaded Simple setup No conflict management required
• Multi-Master Writes can be sent to any of the multiple masters which replicate
them to other masters and slaves Conflict Management required Deadlocks possible if same data is simultaneously modified at
multiple places
Page 26
2266Creative Commons Sharealike Attributions Noncommercial
Replication Considerations• Asynchronous
Guaranteed, but out-of-band replication from Master to Slave Master updates its own db and returns a response to client Replication from Master to Slave takes place asynchronously Faster response to a client Slave data is marginally behind the Master Requires modification to App to send critical reads and writes to
master, and load balance all other reads
• Synchronous Guaranteed, in-band replication from Master to Slave Master updates its own db, and confirms all slaves have updated
their db before returning a response to client Slower response to a client Slaves have the same data as the Master at all times Requires modification to App to send writes to master and load
balance all reads
Page 27
2277Creative Commons Sharealike Attributions Noncommercial
Replication Considerations
• Replication at RDBMS level Support may exists in RDBMS or through 3rd party tool Faster and more reliable App must send writes to Master, reads to any db and critical reads
to Master
• Replication at Driver / DAO level Driver / DAO layer ensures
• writes are performed on all connected DBs
• Reads are load balanced
• Critical reads are sent to a Master
In most cases RDBMS agnostic Slower and in some cases less reliable
Page 28
2288Creative Commons Sharealike Attributions Noncommercial
Real Application Cluster
• All DB Servers in the cluster share a common storage area on a SAN
• All DB servers mount the same block device
• The filesystem must be a clustered file system (eg GFS / OFS)
• Currently only supported by Oracle Real Application Cluster
• Can be very expensive (licensing fees)
DBServer
SAN
DBServer DBServer
Database
Page 29
2299Creative Commons Sharealike Attributions Noncommercial
Recommendation
• Try and choose a DB which natively supports Master-Slave replication
• Use Master-Slave Async replication
• Write your DAO layer to ensure writes are sent to a single DB reads are load balanced Critical reads are sent to a
master
DBServer
DBServer
DBServer
Load Balanced App Servers
Writes & Critical Reads
Other Reads
Page 30
3300Creative Commons Sharealike Attributions Noncommercial
Step 6 – Horizontal Scaling (DB)
• Our architecture now looks like this
• Positives As Web servers grow, Database
nodes can be added DB Server is no longer SPOF
• Negatives Finite limit
Load Balanced App Servers
DB Cluster
DB DB DB
SAN
Page 31
3311Creative Commons Sharealike Attributions Noncommercial
Step 6 – Horizontal Scaling (DB)
• Shared nothing clusters have a finite scaling limit Reads to Writes – 2:1 So 8 Reads => 4 writes 2 DBs
• Per db – 4 reads and 4 writes 4 DBs
• Per db – 2 reads and 4 writes 8 DBs
• Per db – 1 read and 4 writes
• At some point adding another node brings in negligible incremental benefit
Read
sW
rites
DB1 DB2
Page 32
3322Creative Commons Sharealike Attributions Noncommercial
Step 7 – Vertical / Horizontal Partitioning (DB)
• Introduction Increasing the number of DB
Clusters by dividing the data
• Options Vertical Partitioning - Dividing
tables / columns Horizontal Partitioning - Dividing by
rows (value)
Load Balanced App Servers
DB Cluster
DB DB DB
SAN
Page 33
3333Creative Commons Sharealike Attributions Noncommercial
Vertical Partitioning (DB)
• Take a set of tables and move them onto another DB Eg in a social network - the users
table and the friends table can be on separate DB clusters
• Each DB Cluster has different tables
• Application code or DAO / Driver code or a proxy knows where a given table is and directs queries to the appropriate DB
• Can also be done at a column level by moving a set of columns into a separate table
App Cluster
DB Cluster 1
Table 1Table 2
DB Cluster 2
Table 3Table 4
Page 34
3344Creative Commons Sharealike Attributions Noncommercial
Vertical Partitioning (DB)
• Negatives One cannot perform SQL joins or
maintain referential integrity (referential integrity is as such over-rated)
Finite Limit
App Cluster
DB Cluster 1
Table 1Table 2
DB Cluster 2
Table 3Table 4
Page 35
3355Creative Commons Sharealike Attributions Noncommercial
Horizontal Partitioning (DB)
• Take a set of rows and move them onto another DB Eg in a social network – each DB
Cluster can contain all data for 1 million users
• Each DB Cluster has identical tables
• Application code or DAO / Driver code or a proxy knows where a given row is and directs queries to the appropriate DB
• Negatives SQL unions for search type queries
must be performed within code
App Cluster
DB Cluster 1
Table 1Table 2Table 3Table 4
DB Cluster 2
Table 1Table 2Table 3Table 4
1 million users 1 million users
Page 36
3366Creative Commons Sharealike Attributions Noncommercial
Horizontal Partitioning (DB)
• Techniques FCFS
• 1st million users are stored on cluster 1 and the next on cluster 2 Round Robin Least Used (Balanced)
• Each time a new user is added, a DB cluster with the least users is chosen
Hash based• A hashing function is used to determine the DB Cluster in which the
user data should be inserted Value Based
• User ids 1 to 1 million stored in cluster 1 OR
• all users with names starting from A-M on cluster 1 Except for Hash and Value based all other techniques also require
an independent lookup map – mapping user to Database Cluster This map itself will be stored on a separate DB (which may further
need to be replicated)
Page 37
3377Creative Commons Sharealike Attributions Noncommercial
Step 7 – Vertical / Horizontal Partitioning (DB)
Load Balanced App Servers
DB Cluster
DB DB DB
DB Cluster
DB DB DB
LookupMap
SAN
• Our architecture now looks like this
• Positives As App servers grow, Database
Clusters can be added
• Note: This is not the same as table partitioning provided by the db (eg MSSQL)
• We may actually want to further segregate these into Sets, each serving a collection of users (refer next slide
Page 38
3388Creative Commons Sharealike Attributions Noncommercial
Step 8 – Separating Sets
Load Balanced App Servers
DB Cluster
DB DB DB
DB Cluster
DB DB DB
LookupMap
SAN
Load Balanced App Servers
DB Cluster
DB DB DB
DB Cluster
DB DB DB
LookupMap
SAN
Global RedirectorGlobalLookup
Map
SET 1 – 10 million users SET 2 – 10 million users
• Now we consider each deployment as a single Set serving a collection of users
Page 39
3399Creative Commons Sharealike Attributions Noncommercial
Creating Sets
• The goal behind creating sets is easier manageability• Each Set is independent and handles transactions for a
set of users• Each Set is architecturally identical to the other• Each Set contains the entire application with all its data
structures• Sets can even be deployed in separate datacenters• Users may even be added to a Set that is closer to them
in terms of network latency
Page 40
4400Creative Commons Sharealike Attributions Noncommercial
Step 8 – Horizontal Partitioning (Sets)
App ServersCluster
DB Cluster
SAN
Global Redirector
SET 1
DB Cluster
App ServersCluster
DB Cluster
SAN
SET 2
DB Cluster
• Our architecture now looks like this
• Positives Infinite Scalability
• Negatives Aggregation of data across sets
is complex Users may need to be moved
across Sets if sizing is improper Global App settings and
preferences need to be replicated across Sets
Page 41
4411Creative Commons Sharealike Attributions Noncommercial
Step 9 – Caching
• Add caches within App Server Object Cache Session Cache (especially if you are using a Central Session
Store) API cache Page cache
• Software Memcached Teracotta (Java only) Coherence (commercial expensive data grid by Oracle)
Page 42
4422Creative Commons Sharealike Attributions Noncommercial
Step 10 – HTTP Accelerator
• If your app is a web app you should add an HTTP Accelerator or a Reverse Proxy
• A good HTTP Accelerator / Reverse proxy performs the following – Redirect static content requests to a lighter HTTP server (lighttpd) Cache content based on rules (with granular invalidation support) Use Async NIO on the user side Maintain a limited pool of Keep-alive connections to the App Server Intelligent load balancing
• Solutions Nginx (HTTP / IMAP) Perlbal Hardware accelerators plus Load Balancers
Page 43
4433Creative Commons Sharealike Attributions Noncommercial
Step 11 – Other cool stuff
• CDNs• IP Anycasting• Async Nonblocking IO (for all Network Servers)• If possible - Async Nonblocking IO for disk• Incorporate multi-layer caching strategy where required
L1 cache – in-process with App Server L2 cache – across network boundary L3 cache – on disk
• Grid computing Java – GridGain Erlang – natively built in
Page 44
4444Creative Commons Sharealike Attributions Noncommercial
Platform Selection Considerations
• Programming Languages and Frameworks Dynamic languages are slower than static languages Compiled code runs faster than interpreted code -> use
accelerators or pre-compilers Frameworks that provide Dependency Injections, Reflection,
Annotations have a marginal performance impact ORMs hide DB querying which can in some cases result in poor
query performance due to non-optimized querying
• RDBMS MySQL, MSSQL and Oracle support native replication Postgres supports replication through 3rd party software (Slony) Oracle supports Real Application Clustering MySQL uses locking and arbitration, while Postgres/Oracle use
MVCC (MSSQL just recently introduced MVCC)
• Cache Teracotta vs memcached vs Coherence
Page 45
4455Creative Commons Sharealike Attributions Noncommercial
Tips
• All the techniques we learnt today can be applied in any order
• Try and incorporate Horizontal DB partitioning by value from the beginning into your design
• Loosely couple all modules• Implement a REST-ful framework for easier caching• Perform application sizing ongoingly to ensure optimal
utilization of hardware
Page 46
Intelligent People. Uncommon Ideas.
4466
Questions??
[email protected]
http://directi.comhttp://careers.directi.com
Download slides: http://wiki.directi.com