Azure database

Oracle on Azure: Managed Service vs Managed Storage

[Azure, Cloud Volumes ONTAP, Database, Advanced, 8 minute read, Azure database]

Oracle is the preferred RDBMS solution for line-of-business applications, owing to its proven performance, data integrity, security, and portability features. Deploying Oracle on Azure has become increasingly popular due to the close collaboration and commitment of Oracle and Microsoft to deliver and end-to-end support for Oracle-dependent enterprise applications with Azure database workloads.

But there’s more than one way to deploy Oracle on Azure: using a managed database service option by integrating with Oracle Cloud Infrastructure (OCI), or deploying in a virtual instance from the Azure Marketplace using managed storage. Both of these solutions are certified by Oracle to be deployed in Azure, with out-of-the-box lifecycle support from Microsoft. Which is right for you?

This blog will explore both options for deploying Oracle on Azure, delve into the pros and cons and best fit solution for enterprise scenarios, and see how Cloud Volumes ONTAP for Azure can help.

Oracle: Introduction to Deployment in Azure

Oracle is a multi-model relational database solution, best suited for large-scale enterprise clustered applications like ERP and data warehousing. It has been a popular solution for years for on-premises deployments. With cloud adoption on the rise, the focus has now shifted to the optimal migration and deployment approach of this trusted database platform in the cloud. Oracle is available in various editions, of which the Enterprise and Standard editions are certified by Oracle to be deployed on Azure. Let’s look into some of the key considerations while planning for Oracle deployment in Azure.

Deployment models: One popular option to deploy an Oracle database on Azure is to use the database-as-a-service (DBaaS) model. In a DBaaS, the platform-level configuration and management are done by Azure and the customer owns the database specific configuration. Unlike other DB platforms such as MSSQL and MySQL, Oracle does not have its own DBaaS offering in Azure at the time of this writing. Instead, Microsoft in collaboration with Oracle offers integration of Azure infrastructure with Oracle Cloud Infrastructure (OCI) so that Oracle applications deployed in Azure can seamlessly connect to backend databases.

High bandwidth and secure connection between the OCI database layer and the Oracle application layer in Azure can be enabled using Azure ExpressRoute and Oracle FastConnect. Many of the popular Oracle application stacks including E-Business suite, PeopleSoft, Oracle Retail applications and more are certified by Oracle to be deployed using this OCI interconnect model.

For customers who want full control of the deployment stack, starting from OS and up to the application layer, Azure supports deployment of Oracle in an infrastructure-as-a-service (IaaS) model using Oracle Linux images in the Azure Marketplace. Customers can also choose to bring their own image to Azure or deploy the database independently in a non-Oracle Linux image from the Marketplace. Azure also supports popular Oracle applications such as Oracle WebLogic as readily available Marketplace images integrated with the Azure Application Gateway and Azure Active Directory.

Licensing: Oracle offers license mobility, which allows customers to bring their existing licenses and use them for Oracle applications deployed in Azure VMs. This bring-your-own approach for Azure licenses provides greater flexibility and portability for organizations that have heavily invested in the Oracle suite of products for on-premises usage. By using a BYOL, cloud adoption is much easier, as they can easily spin up the instances in Azure using a pre-configured image and reuse the licenses.

OCI database services support both bring-your-own-license as well as pay-as-you-go options. The BYOL option in OCI supports all advanced database functionalities including backup/restore, patch updates, and Oracle Data Guard. You could also switch the database licensing type after provisioning of the database, which offers additional flexibility in terms of license management.

Data migration: The ease with which data can be migrated to cloud plays an important role in the success of any digital transformation program. Migrating Oracle databases to Azure can be achieved using a number of methods with the IaaS model, the simplest of them being performing an RMAN backup  and restoring to Oracle VM in Azure. Other tools that are commonly used for data migration are Oracle Data Guard and GoldenGate. Data migration to OCI can also be done using OCI Storage Gateway service and RMAN backup.

OCI-Managed Service for Azure and Oracle Integration

Oracle Autonomous Database is a DBaaS service offering from Oracle Cloud Infrastructure that can be integrated with your Oracle applications deployed in Azure to deliver a full stack of Oracle services. Autonomous Database supports transaction processing and data warehousing workloads. Since Autonomous Database is delivered as a managed service, the routine maintenance activities like backup, upgrades, patching, and database tuning are all handled by the OCI platform.

Oracle Autonomous Database supports Oracle database 18c and 19c; however, these two versions are not available in all regions. This might pose a limitation to customers based on their geography, even if the connected Azure infrastructure is able to support the application layer.

For OCI and Azure to work together, an Azure ExpressRoute circuit should be integrated with OCI FastConnect service. The prerequisite for this configuration is that the peering location of both the services should be in proximity to each other. It should also be mentioned here that this interconnection is supported in only a few Azure regions.

Autonomous Database in OCI makes it possible to automate the creation of full backups on a weekly basis, followed by daily incremental backups. However, the retention period of those backups is limited to 60 days. The automated schedule of backup may not work well with the organizational requirements and hence customers are advised to look into additional backup configuration options, either manually or using OCI SDK and Rest API calls to initiate the backup programmatically.

Since OCI supports only the latest version of Oracle, organizations with dependency on legacy versions would have to make additional investments to either upgrade their database or modernize their applications to suit the available database version. Limited version availability of Oracle, availability of database versions in a given geography, and regional restrictions of cross-region connectivity between OCI and Azure should all be taken into consideration while adopting the managed service.

Oracle in Azure IaaS: Managed Storage Instances

The IaaS deployment model that uses Azure compute and storage services offer a more flexible alternative to the DBaaS deployment option of Oracle. Customers can leverage the custom-built VM SKUs and employ additional managed storage services tuned for IOPS intensive database operations in Azure. The IOPS available for Oracle databases can be greatly improved by distributing the data across multiple disks managed by Oracle Automated Storage Management (ASM). The collective IOPS from the disks deliver much improved storage performance and faster database transactions. Choosing the right storage management solution makes a world of difference here.

Azure managed disks, the managed storage service from Azure, offers different performance SKUs such as premium and ultra disks, which can be used for deploying Oracle databases in the IaaS model. The bring-your-own-license option considerably reduces the operational costs and enables customers to reuse their existing licenses. However, it should be noted that compared to a DBaaS solution, there are additional responsibilities to be addressed here: i.e., factors such as high availability, backup, patching, upgrades, failovers, etc., especially in production deployments. Customers can consider specialized managed storage solutions like NetApp Cloud Volumes ONTAP, which deliver advanced data management capabilities when compared to native cloud storage solutions to address these requirements of production environments.

An Oracle with Managed Storage Case Study

One company that chose to deploy an Oracle database in the IaaS model is a major online fashion retailer with headquarters in the United Kingdom. The company offers a selection of hundreds of brands as well as its own line of clothing and accessories and ships to countries worldwide. This company’s entire retail stack is based on Oracle, making it the definition of mission-critical. Using the managed storage option with Cloud Volumes ONTAP, this company was able to ensure the database has high availability across regions to ensure their business continuity, without the restrictions of the DBaaS model.

Conclusion: Choosing the Best Model for Your Database

While DBaaS Oracle options in Azure using OCI ease the infrastructure deployment requirements, there are many additional challenges associated with them: restricted versioning, regional availability, and integration complexity with Azure being the most prominent. Deployment in the IaaS model using managed storage solutions such as Cloud Volumes ONTAP can help ease up these concerns, as this model provides end-to-end flexibility for the deployment with any Oracle database version in any Azure region.

Cloud Volumes ONTAP augments the native Azure cloud storage layer by delivering an enterprise-class data management solution. These benefits include:

  • Full control over how databases are deployed and managed in Azure
  • Data hybridity and mobility to make it easy to keep the data where you want it, even in different clouds or on-prem
  • Scalability and advanced features not available with DBaaS approach
  • No restrictions on Oracle versions or Azure regions
  • Built-in intelligent caching to decrease latency and increase database performance
  • Storage efficiency technology that lowers the overall cost of the database
  • Cost-efficient snapshot and cloning capabilities
  • A high availability configuration option to ensure RPO=0 and RTO < 60 seconds
  • Easy migration from existing NetApp appliances with SnapMirror® or from elsewhere with Cloud Sync

Ultimately, the choice between the two models will come down to which is the best fit for your current database and business needs. While cloud native storage solutions could restrict you to specific cloud service provider environments, Cloud Volumes ONTAP enables data hybridity where the data can be migrated on the fly across different cloud platforms from your on-premises environment. The NetApp SnapMirror data replication feature of Cloud Volumes ONTAP helps you with data migrations, which is often a pain point for all cloud adoption initiatives. Unlike a DBaaS service where additional backup requirements are to be addressed manually, the built-in snapshot feature of Cloud Volumes ONTAP enables protection for your Oracle databases with minimal configuration overhead.

An Oracle database is a critical component of your business. Make the right choice when deploying it in the cloud.

New call-to-action