Running Siebel on AWS - Oracle Open World 13
Download as PPT, PDF5 likes2,750 views
This document discusses deploying Siebel CRM on AWS to provide a highly available and scalable architecture with no hardware requirements. Key points include: - Enterprise Beacon has expertise in both Siebel and AWS and launched the first Siebel test drive on AWS. - Deploying Siebel on AWS provides benefits over traditional on-premise or SAAS models like lower costs, built-in high availability, disaster recovery, unlimited environments, and AWS service level agreements. - The proposed Siebel on AWS architecture spreads application tiers across multiple availability zones for high availability and uses services like Elastic Load Balancing, Route 53, and database multi-AZ for automatic failover and disaster recovery across regions.
Running Siebel on AWS - Oracle Open World 13
- 1. Siebel on AWS Building extremely HA and scalable Siebel infrastructure with no Hardware! Enterprise Beacon, Inc. [email protected]
- 2. Enterprise Beacon, Inc. • Founded in 2008 • Oracle Gold Partner • AWS Consulting Partner • Launched the first test drive for Siebel using EC2 and RDS – Deep expertise on both Siebel and AWS • Object Hive - Exa-Ready Certified – Siebel Version Control Product
- 3. Benefits over Salesforce/SAAS Siebel on AWS Salesforce/SAAS Customers already own Siebel licenses! Only pay for AWS based on usage. Per user/Per month = High TCO No high cost/high risk reimplementation needed for existing Siebel implementations High risk/ High cost migration Ability to completely keep instance behind enterprise firewall Has to be accessed over public internet No crippling governor limits in Governor and edition limits on
- 4. Benefits over Salesforce/SAAS.. Siebel on AWS Salesforce/SAAS Apply Innovation packs as needed Quarterly releases are applied directly Things break: e.g Sidebar bar JS hack. History data storage durations. No instance migrations Customers can be moved across instances without a choice with little notice Unlimited number of environments Limitation on number/size of environments and refresh periods AWS 99.95 guarantee No SLA guarantee!
- 6. Benefits over On Premise AWS On Premise Low operating costs and TCO Capital Expense and Higher TCO for hardware and virtualization licenses. Moore's law works for you Moore's law works against you Ability to right size any environment when needed. Not possible Auto recovery for failed instances. Not possible. Even with virtualization you still need spare hardware. HA built into the platform can be taken advantage of by spreading instances across AZs and Regions HA requires expensive clustering or expensive virtualization software and hardware. Does not protect against datacenter failure (No Azs)
- 7. Benefits over On Premise .. AWS On Premise DR can be built at extremely low cost across data centers DR requires buying expensive hardware Low Support cost High Hardware support costs VMware licensing and support costs Ability to know the performance before launching Performance is known until hardware is procured and deployed No End Of Life for hardware. Simply upgrade to newer instances with a few clicks Hardware EOL customers are forced to buy hardware again Ease of use. Create replica environments in few clicks Hard to maintain and replicate
- 8. Cost to Business • All this leads to – Wasted capacity – Inadequate capacity – High Costs – Inadequate testing coverage – Poor testing – Poor build and release quality
- 9. AWS Concepts • EC2 (Elastic Compute 2) – Amazon Elastic Compute Cloud (EC2) is a central part of Amazon Web Services (AWS). EC2 allows users to rent virtual computers on which to run their own computer applications. – Many different instance types • RDS (Relational Database Service) – IAAS more specifically Database As A Service – Available database options for Siebel • Oracle • SQL server • Elastic ComputeUnit (ECU) – The elastic compute unit (ECU) was introduced by Amazon EC2 as an abstraction of computer resources and as a unit to measure computing capacity
- 10. • Elastic Block Store (EBS) – All virtualized instances run of a stored image – EBS volumes are built on replicated storage, so that the failure of a single component will not cause data loss. • Simple Storage Service (S3) – Amazon S3 provides a simple interface that can be used to store and retrieve any amount of data, at any time, from anywhere. It gives customers access to the same highly scalable, reliable, secure, fast, inexpensive infrastructure. • Elastic Load Balancer (ELB) – In addition to balancing load based on incoming traffic ELB can detect unhealthy instances and re-route traffic to healthy instances • Route 53 – A highly available and scalable Domain Name System (DNS) web service, it can help detect an outage of a instance and redirect traffic to alternate locations where your application is operating properly. – 100% SLA for Route 53 since it used build DNS Fail Overs
- 11. North America
- 12. Gov Cloud For US Gov Only • AWS Region designed to allow US government agencies and customers to move sensitive workloads into the cloud by addressing their specific regulatory and compliance requirements • FEDRAMP and ITAR certified
- 13. South America
- 14. EMEA
- 15. Asia Pacific
- 16. Availability Zone A data center build for the cloud • Each region is further divided into Zones called Availability Zones (AZs) • To make AWS more fault-tolerant, Amazon engineered Availability Zones that are designed to be insulated from failures in other availability zones. – Separate Power Sources – Separate Switches – Separate storage – Separate physical infrastructure • AWS SLA gurantee of 99.95 for cross AZ
- 17. EC2 • Various instance sizes and types – From 1 core/1gb to 40Core/244gb • Easily change instance size (ECU/Memory) with web console or AWS APIs • Easily backup and restore images (called AMIs) – `
- 18. EC2.. • Provisioned IOPS – tailor storage performance to your need • Only storage charges apply for shutdown instances – $0.05 per GB per month – So 120gb stand by instance will cost only $6 a month. • All web console task have equivalent APIs • Auto Recovery
- 19. AWS pricing model • On-Demand – pay for compute capacity by the hour with no long-term commitments. • Reserved instance – option to make a low, one-time payment for each instance you want to reserve and in turn receive a significant discount on the hourly charge for that instance.
- 20. RDS • Amazons database as a service for Oracle and SQL Server. • Automatic Backups for user define period • Point in time recovery • Provisioned IOPS currently up to 30000 IOPS for extreme IO performance • Multi – AZ option • BYOL/Pay as you go for Oracle and MS SQL
- 21. VPC Virtual Private Cloud • A virtual private cloud (VPC) is a virtual network dedicated to your AWS account. It is logically isolated from other virtual networks in the AWS cloud. • Three connectivity options between customers data center and AWS’s data center – Software Encryption – Hardware Encryption – Direct connect establishes private connectivity between AWS and your datacenter, office, or collocation environment.
- 22. Subnets Border Gateway Protocol • A subnet is a range of IP addresses in your VPC • These address are unique private IP addresses across customers data center • Users need not know the addresses are AWS or no • Instances on AWS can be attached to customers domain
- 23. Siebel Architecture on AWS Enabling Siebel on the cloud
- 25. Solution Overview • Customers network can be connected by • IPSec/VPN or Hardware/VPN or Direct Connect • All instances will only be accessible from customers network only and part of customers domain • Each application tier will be placed across multiple AZs • Standby images will be created for Disaster recovery in a Alternate Zones/ Regions • A Elastic Load balancer, Elastic IP and Route 53 will be used to make failure over transparent to end users.
- 26. Siebel Web Servers on AWS
- 27. Siebel App Servers on AWS
- 28. Siebel Gateway on AWS
- 29. Database on AWS • In the region closet to the customer site deploy a Multi-AZ RDS instance • In a secondary region have a stand by – Will require Golden gate or other means • Failover to the stand by using Route 53
- 30. Unbeatable combination. Win! Win! • Open UI (Mobile First) – Mobility • Innovation Packs (No Major Upgrades) – Pick and Pace your upgrades • AWS (Cloud First) – On Demand, HA, unlimited scalability , Low Op Ex and TCO • Enterprise Beacon – Outcome Management
- 31. Demos
- 32. Paradigm Shift No more hardware! • Customers need to get out of the old paradigm – Cloud is not just Physical to Virtual – Take advantage of underlying HA – Replicate full stack when needed – Fine tune instance types – Auto scale – Automatic Backups
- 33. AWS Support • Subject to AWS approval – 50% discount for POC for customers – $25,000 credit for 1st year of operation for 1st reference able customer
- 34. Enterprise Beacon Roadmap • Migrate non prod environments < 1 week • Migrate prod environments – Define architecture based on expected SLA – Build Infrastructure – Performance test and tune infrastructure – Help redefine ops process – Cut Over – Operations support
- 35. Next Steps?
- 36. Appendix • Web Server failure scenarios • App Server failure scenarios • Gateway failure scenarios • Database failure scenarios
- 37. Scenario Solution Impact Web Server Fails Auto recovery Slow Performance while stand by instances are added AZ failure ELB Managed. Launch Additional Instances on available AZ Slow Performance while stand by instances are added Region Failure * i.e failure of all AZs in a region which is extremely low possibility DNS Failover. Startup web servers on the secondary region. Switch Route 53 entry for Web Server to point to ELB in Secondary Region. System will be unavailable while instances in stand by region come online •Impact can reduced if instances in secondary region are kept online High Load/Low Load Based on concurrent users, CPU and Memory additional standby instances can be turned on or existing Improved performance and TCO as infrastructure scales to according to demand
- 38. Scenario Solution Impact App Server Fails Auto Recovery Reduced performance till instance is replaced AZ Failure Startup additional instance in online AZs Reduced performance till instance is replaced Region Failure Startup app servers on the secondary region. System will be unavailable while instances in stand by region come online * Impact can reduced if instances in secondary region are kept online High Load/Low Load Based on concurrent users, CPU and Memory additional standby app servers can be turned on or existing app servers can be turned off Improved performance and TCO as infrastructure scales to according to demand
- 39. Scenario Solution Impact Gateway Server Machine fails Autorecovery User sessions continue to run as is. Siebel Servers can not be started/shutdown and Siebel Admin disabled until secondary AZ gateway is online. AZ Failure Copy the siebns.dat from S3 bucket to secondary Gateway. Start standby in secondary AZ. User sessions continue to run as is. Siebel Servers can not be started/shutdown and Siebel Admin disabled until secondary AZ gateway is online. Region Failure Copy the siebns.dat from S3 bucket to standby instance in different region. Update Route 53 to point to secondary region gateway User sessions continue to run as is. Siebel Servers can not be started/shutdown and Siebel Admin disabled until secondary gateway form another region is online.
- 40. Scenario Solution Impact Active RDS instance fails RDS fails over to secondary node and that becomes primary System is down till the secondary node comes online Region Failure Golden Gate Replication
- 41. Thank You