A leading software-as-a-service (SaaS) vendor that drives critical aspects of businesses, from supply chain management and autonomous commerce to procurement and vendor management, places high availability (HA) as a top priority.
“If our client’s software goes down, their customers lose money, so it’s vital to have services running 24/7. As a result, maintenance windows are extremely narrow, and systems can’t be shut down to install software updates.”
Art Danielov, CEO of FlashGrid
Migrate Oracle Database to AWS and upgrade to Oracle Database 19c SE HA.
Achieve high availability and enable high DBA productivity.
FlashGrid Cluster for Oracle Failover HA on AWS.
To fully embrace the benefits of cloud-native software delivery, the SaaS vendor decided to migrate the company’s databases from its on-premises data center to a public cloud, specifically Amazon Web Services (AWS). The DBA team carefully evaluated options to run Oracle databases on AWS, including DBaaS (database-as-service) solutions such as Amazon RDS for Oracle. However, based on what they found, they continued exploring self-managed options.
Before moving to AWS, the company delivered its software services from an on-premises data center using Oracle Database 12c Standard Edition 2 and Oracle Real Applications Clusters (RAC). Migrating to the cloud offered an opportunity to upgrade to Oracle Database 19c without additional downtime. Because RAC is unavailable in Oracle Database 19c Standard Edition 2, the team chose Standard Edition High Availability (SEHA) to provide the necessary fault tolerance capability and prevent outages.
Making all these changes together posed a unique challenge. The Oracle DBA team was used to having full control of everything from disk to network. They needed to be able to perform tasks such as installing third-party drivers and connectors to ship data to the SQL server and perform extract, transform, and load (ETL) jobs to Snowflake.
To ensure a reasonable learning curve for his team, the lead DBA needed a solution that would simplify database management in the cloud and provide the flexibility needed to work in heterogeneous database environments and respond to changing requirements. He would try to reduce the need to retrain and its expense as much as possible.
The SaaS provider’s AWS representative suggested speaking with FlashGrid. The lead DBA quickly realized that with FlashGrid Cluster on AWS, his team would keep full control and use familiar Oracle tools such as ASM and Clusterware. He perceived FlashGrid Cluster as a layer underneath their existing stack that makes it easier to use. Unlike DBaaS, FlashGrid allowed them to use all the Oracle features and keep the level of control necessary to do their jobs without imposing restrictions on the functionality and interoperability of the database system.
One downside of DBaaS architecture was the need for more flexibility to optimize storage costs. When database storage exceeds 100 TB, it becomes a significant component of the infrastructure’s total cost. FlashGrid Cluster diversifies storage options, making the AWS environment more cost-efficient.
With FlashGrid, the AWS user takes full advantage of tiered storage with allocations managed by the DBA team. Ninety percent of their data is in Large Object (LOB) tablespaces; FlashGrid allows them to employ SC1 volumes, the lowest tier of Amazon EBS storage, instead of more expensive GP3 volumes. The savings are substantial because SC1 storage is less than 20% of the cost of GP3 storage. The company still uses GP3 volumes for transactional parts of the database, where performance is paramount.
The SaaS vendor’s migration to AWS involved many moving parts and new technologies, which pressured the lead DBA to stay on schedule and resolve issues quickly. Fortunately, FlashGrid was more than up to the task; FlashGrid experts answered even the most technical questions down to the level of the Linux operating systems without exception.
“One factor contributing to the customer’s positive experience is that their technical people could connect directly with technical counterparts at FlashGrid – expert to expert. Most of their questions were answered in a single call.”
Art Danielov, CEO of FlashGrid
The bottom line is that FlashGrid delivered on its promise to its customer. The DBA team ran HA tests, and the systems behaved as expected. The application seamlessly connected to the surviving instance/node during failover without any problems. Since the cutover from Oracle Database 12c running on-premises with RAC to Oracle Database 19c SEHA running in the AWS cloud, the SaaS company has not experienced a single outage.
It’s hard to measure peace of mind. Still, there’s little doubt that FlashGrid contributed to the migration’s success and instilled confidence in the customer’s DBA team. The lead DBA shared, “We are pleased that FlashGrid allowed us to retain all the key features from our on-premises environment and streamlined the migration and upgrade.”
To this day, the SaaS provider continues to use FlashGrid Cluster for Oracle Failover HA to run Oracle 19c SEHA on AWS and has maintained perfect uptime.
We will contact you shortly.
Try again later.
