1. Auto-scaling Axis2 Web ServicesonAmazon EC2 By Afkham Azeez (azeez@apache.org) WSO2 Inc.

2.

The Problem

A Solution

Some Concepts

Design & Implementation Details

• Overview

3.

Fault tolerance, high availability & scalability areessential prerequisites for any enterprise applicationdeployment

One of the major concerns is avoiding single points offailure

There is a high cost associated with achieving highavailability & scalability

Need to achieve high availability & scalability atanoptimum cost

• The Problem

4.

Traditional solution

• Buying safety-net capacity

Better solution

• Scale-up the system when the load increases

• Scale-down the system when the load decreases

• Should not have idling nodes

• Pay only for the actual computing power &bandwidth utilized

• Solutions

5.

Building a framework which will auto-scale thenumber of Axis2 nodes on Amazon EC2,depending on the load

• Project Objective

6.

• Axis2 is a middleware platform which enableshosting of Web service applications and supportssome of the major Web services standards

• Can host Web services written in Java as well asvarious scripting languages

• Can be deployed in a clustered configuration

• Uses Apache Tribes for clustering

• Axis2 clustering has been adopted to work on EC2

• Apache Axis2

7.

• Apache Synapse is designed to be a simple,lightweight and high performance ESB

• Supports load balancing with or without failover

• Supports static & dynamic load balancing

• Uses Apache Axis2

• Apache Synapse

8.

• A messaging framework with group communicationabilities

• Allows you to send and receive messages over anetwork, it also allows for dynamic discovery ofother nodes in the network.

• Used by Apache Tomcat & Apache Axis2

• Apache Tribes

9. Deploying a Service on the Cloud AMI Instances 10. Deploying a Service on the Cloud 11. Deploying a Service on the Cloud 12. Deploying a Service on the Cloud Service is available now 13. Auto-scaling 14. Auto-scaling Load Increases 15. Auto-scaling Startup new instances 16. Auto-scaling New instances join group 17. Auto-scaling Load Decreases 18. Auto-scaling Terminate instances 19. Deployment Architecture 20. Membership Aware DynamicLoad Balancing 21. Membership Aware DynamicLoad Balancing 22. Membership Schemes

Static

Dynamic

Hybrid (WKA based)

23. WKA Based Membership (1/3) Application member joins. The load balancer is also a well-known member 24. WKA Based Membership (2/3) A non-WK load balancer joins 25. WKA Based Membership (3/3) A well-known load balancer rejoins after crashing 26. Membership Channel Architecture 27. Initialization Channel Architecture 28. Synapse Configuration - axis2.xml 29. Synapse Configuration - synapse.xml 30. Synapse Configuration

AutoscaleInMediator

AutoscaleOutMediator

Autoscale Task

31. Synapse ConfigurationNormal Flow 32. Synapse ConfigurationFault Flow 33. Synapse Autoscale Task 34. Synapse Autoscale Task

Sanity Check

Autoscaling

• Scale up

• Scale down

35. Booting up Start Start Start Start Initial Instance LB Group Axis2 App Group synapse.xml S3 Bucket Load configuration 36. Axis2 Configuration - axis2.xml 37. More Implementation Details

Single AMI autoscalews

Start

• ec2-run-instances ami-a03fdbc9 -k autoscale -f payload.zip -gautoscale-lb

Payload

• Extract params from payload

• Env variables

Axis2 & Synapse Configuration files & Repositories

• Maintained on S3

Fault Tolerance

• Monitoring cron job

• Java Service Wrapper daemons

• Future: Use monit

• Future: Axis2/Synapse agent to check process status

38. Axis2 Configuration & Repository Repo Conf Axis2 Instance S3 Bucket Load repo Load configuration 39. Questions Suggestions Improvements 40. Thank You