Many organizations have business critical workloads that rely on Oracle and they are on a journey to modernize these applications. They recognize the importance of still locating their Oracle workloads in tried and tested environments to reduce risk to their business.
Re-hosting approach is the best fit for migration from current hosting provider to a new infrastructure provider. As retail applications are more complex with a larger number of components with dependencies, then this approach can be quite challenging, especially when minimal migration down-time is a business requirement.
Oracle Retail had several challenges like compatibility, architecture difference, Lift and Shift limitations. These have been simplified by optimizing the approach, utilizing best of breed solution of public cloud provider.
• Achieving High Availability
• Matching EXADATA performance
• Optimizing Environment build time
• Reducing downtime during cutover
• Other unknowns, related to functionality
The objective of this document is to provide a technical overview of the solution, approach to overcome with challenge and non-functional aspects of the solution like High availability, Disaster Recovery, Backup, Monitoring, Automation and Security.
- Cloud Setup:
Build the Azure landing zone subscription and VNETs standard model and security standards which includes but not limited to:
- Network Components such as VNET peering with existing ExpressRoute Circuits, Route Tables (UDRs), Firewall burns, Approved traffic flow patterns.
- Security standards such as Role based access control (RBAC), Azure AD Identity and Access management, Azure Defender, Private endpoints, tagging and other cloud security policies.
- Effective logging and monitoring components for Azure environment.
- Building VM OEL images with appropriate agents such as CrowdStrike, New Relic and Commvault. This can be changed as per customer requirement.
- Achieving High Availability and Disaster Recovery:
· Database: Oracle Active Data Guard is the comprehensive solution available to eliminate single points of failure for critical Oracle databases. It prevents data loss and downtime in the simplest and most economical manner by maintaining a synchronized physical replica of a production database at a remote location. If the production database is unavailable for any reason, client connections can quickly, and with some configurations transparently, failover to the synchronized replica to restore service. Oracle dataguard can be provisioned across the availability zones. The primary will be commissioned in zone 1, the standby in zone 2 and the observer in zone 3. The dataguard can be configured for Maximum availability by enabling synchronous replication between the primary and the standby. This will ensure zero data loss during datacenter outages.
The offers the following from a high availability perspective.
- Zero Data loss for the Oracle Databases during a datacenter outage
- Automatic Failover from primary database to a standby database using Oracle Fast start failover (FSFO)
- Seamless application connectivity to the new primary database using Oracle’s transparent application failover.
- No DNS or other configuration changes required after the failover.
- Microsoft offers a financial backed SLA of 99.99% for the Virtual machines across availability zones.
· Application: The Oracle WebLogic clusters which are part of a WebLogic domain will be deployed across the Azure Availability zones. This will help make sure that the WebLogic server instances which are part of the same cluster are active across two or more datacentres. This increases resiliency against datacenter outages and offers a 100% uptime for the Oracle retail suite applications. The applications are active in all the zones where it is deployed and hence no configuration changes would be required in the event of a failure. Azure Load Balancer is zone-redundant, ensuring high availability across Availability Zones. Oracle Weblogic supports transparent application failover so in the event of a failure where the primary fails and the standby assumes the roles of a primary, the application server will automatically and seamlessly connect to the new primary database. No configuration changes or manual intervention would be required on the application or database to facilitate the service continuity.
As a part of the backup policy, the prod backup snapshots will be replicated to the Australia south east region (DR region) which is 900 Kms away from the primary and persisted in the Commvault library for restoration. In the event of a catastrophic failure where the primary region is not available, a new environment can be stood up in the remote region in few minutes by restoring the snapshots. Oracle Archive logs can be applied on the restored Oracle databases to minimize the data loss.
Commvault can be used for the backup solution of Oracle databases. The Azure native backup solution is also available and offers both streaming and snapshot-based backup of the Oracle databases to the recovery services vault. Considering Oracle Retail, the size of the prod and pre-prod Oracle databases which are more than a TB in size, Commvault’s IntelliSnap for the Oracle database backups.
- Matching EXADATA performance:
It is advisable to analyze and validate the required Bill of Materials. Below approach can be taken.
- Database Workload Analysis
- Application Server Usage Analysis
- Setup to-be production like environment including scheduler
- Performance Test with average and pick load. Define Server Sizing
- Optimizing Environment build time (using Automation):
It is always challenging to build complex environment of Oracle Retail which in general is highly customized. Oracle Retail Product is not recommended to use Lift and Shift Approach. Infosys has developed tool (using Chef) which helps reduce build time by 60%.
- Reducing downtime during cutover:
Oracle Golden Gate solution helps reducing downtime during cutover.
- Initial Copy – RMAN Backup and restore to Azure DC. Enable Data replication and Sync using Oracle Goldengate. Implement indexes and any other schema changes based on Performance Test results.
- During Cutover Window, Switchover to Azure and make it primary (R/W), Application Connectivity and PVT and user cutover.
- Standby will be open in R/W and DDL Change can be implemented outside of cutover window, hence less outage time.
- More flexibility to choose schemas to replicate.
- A licensed option.
- Other unknowns, related to functionality:
In Rehosting approach, it is general understanding that extensive testing is not necessary. However, this is not true. Extensive testing should be planned which should include SIT, OAT, SVT, Dress Rehearsal.
Azure Security Center’s features cover the two broad pillars of cloud security, namely, Cloud Security Posture Management and Cloud Workload Protection. The cloud security posture management is a free offering, and it includes secure score, detection of security misconfigurations in your Azure machines, asset inventory, and more. The cloud workload Protection, also known as Azure Defender, brings in advanced intelligent protection capabilities like securing the management ports of your VMs with just-in-time access and adaptive application controls to create allow lists for what apps should and shouldn’t run on your machines. Azure Defender also includes vulnerability scanning for your virtual machines and container registries at no extra cost.
- Monitoring and Notification:
Azure Monitor will be enabled for the Azure virtual machines. VM insights which is a part of Azure monitor helps monitor the performance and health of your virtual machines and virtual machine scale sets, including their running processes and dependencies on other resources. It can help deliver predictable performance and availability of vital applications by identifying performance bottlenecks and network issues and can also help you understand whether an issue is related to other dependencies.
- Reduce Cost: Drive down the cost to run and operate the platform
- Drive Innovation: Leverage the platform for experimentation for new technology & innovation to solve client’s business problems
- Sustain & growth: Enable the ability to address new marketable products & services
- Customer engagement: Deliver high quality customer experience to show the art of the possible