“Database improvement” by migrating to Amazon Aurora PostgreSQL
Achieved licensing costs reduction, improved performance, operability, and development speed
At Home Co., Ltd., a pioneer in the real estate information service industry that provides vast volumes of property-related information and know-how, is shifting its system infrastructure to the cloud in stages. After migrating multiple applications to Amazon Web Services (AWS), the company also shifted its on-premises Oracle Database Real Application Clusters (RAC) to Amazon Aurora PostgreSQL. As a result of its migrations to AWS, including its database infrastructure shift, At Home has seen improvements in scalability and development speed as well as reductions in licensing costs.
The appeal of AWS is that it allows us to develop services at speed by mixing and matching various cutting-edge services. By utilizing AWS, we hope to continue providing services that anticipate our users’ needs.”
Director, General Manager of Information Systems Department, At Home Co., Ltd.
Migrated applications to deliver services that meet business needs in a short time
At Home provides property-related information and know-how to real-estate companies via its extensive real estate information network. It is one of the largest in Japan, and it offers the Real-Estate Information Website, At Home, and a smartphone app to distribute enormous volumes of housing information to consumers. Osamu Yamamoto, Director and General Manager of the Information Systems Department, talks about changes in the company's IT strategy that plays a key role in providing these services.
“Ten years ago, our hardware was installed in our headquarters building, and then it was moved to the data center. Since then, we have moved forward with a shift to the cloud. By effectively combining on-premises and cloud resources, we were able to rapidly deliver services that anticipate end users’ needs and thus contribute more to businesses, all while maintaining high availability.”
AWS played a vital role in the company’s shift to the cloud. “The first time we used AWS was when scalability became an issue in our real estate property image-delivery system,” recalls Hideaki Sugawara, Head of the Systems Infrastructure Management Office. As the number of properties handled by the company increased, expanding on-premises resources such as database and storage became a problem, it migrated to AWS in 2015.
In terms of cost, At Home used to size on premise IT resources to meet peak season demand, which resulted in waste, at times, of over half of the resources. After moving to the cloud, resources could be rolled out only as required by demand, leading to cost optimization. Moving to AWS was smoother than expected, and as the company recognized anew the benefits of scalability and flexibility in securing resources, it continued migration to AWS for its main services as well.
Database migration from Oracle RAC to Amazon Aurora
In 2020, At Home migrated its consumer-facing At Home Real-Estate Information Website to AWS, upon which it shifted from Oracle RAC to Amazon Aurora PostgreSQL in order to reduce expensive licensing costs and embrace maximum benefits of the cloud. The reason it chose PostgreSQL instead of MySQL was to take advantage of the similar writing for stored procedures between Oracle and PostgreSQL, which allowed it to leverage program conversion tools to cut down on development person-hours. In this case, about 80 percent of At Home’s batch processing programs were written in PL/SQL, and so how quickly and inexpensively they could be moved to the cloud were crucial factors. Yusaku Kurimoto, Head of C Systems Development Office, explains the migration process as follows: “First, while moving our applications to AWS in Phase 1, we made plans to carefully proceed with migrations over a period of roughly 3 years. This included the database batch system migration of Phase 2. There were about 200 PL/SQL applications running in the existing environment, but by using tools such as the AWS SCT (Schema Conversion Tool), we were able to automatically convert 70-80 percent of those to PL/pgSQL stored functions. After building a new database environment with Aurora PostgreSQL from September to December, 2019, and then running parallel systems for several months, the migration was completed in July, 2020.”
In addition, the company took advantage of the AWS Database Migration Service (DMS) in its data migration. During the parallel operation period, differential data was extracted from Oracle RAC, converted, and poured into Aurora PostgreSQL. This reportedly minimized downtime during database switchover.
Migrating saved costs, significantly improved performance, operability, and development speed
Two months after migration to Aurora PostgreSQL, At Home’s Oracle RAC was fully decommissioned and license costs were no longer needed. The hardware is still being used for other purposes, but if it can be removed as part of migration to the cloud, At Home can expect further savings, as costs for data center racks, electricity, and hardware maintenance will become unnecessary. As for performance, queries are processed faster than in Oracle RAC. However, even more significant is the reduction in operational burdens, says Toru Shiraishi, Member of the C Media Development Group. “AWS database services offer a feature called Performance Insights, which makes it easy to visually identify heavy SQLs and the underlying causes. We can assess the status across the entire AWS system with Amazon CloudWatch metrics, and execute all required operational controls from the console. Operation is also possible through APIs, and we are actively implementing operational automation as well. When we need to investigate past data, we can easily restore data of a specific time point using the point-in-time recovery function. These functions were not available in our on-premises environment.”
In addition, Kurimoto appreciates the benefits of increased development speed. “As we migrate further onto AWS, the borders between infrastructure and development have diminished, and this has changed our organizational boundaries as well. Nowadays, more and more we see the developers themselves exploring optimal infrastructures. By combining the right AWS services, the time it takes to implement a new service has been reduced to one tenth or lower (of previous levels).”
Comprehensive AWS support and programs to encourage insourcing helped achieve database migration in a short time frame
This database migration was achieved in a short time thanks in part to support from AWS. “The support team quickly resolves issues arising on AWS. We have kept an Enterprise Support contract since 2019. We are provided with straightforward weekly updates by the Technical Account Manager, and are able to consult with solution architects regarding things like how to deploy a new system. At the time of our Oracle RAC data migration, we had our Technical Account Manager join us in creating and reviewing the migration strategy,” says Shiraishi.
At Home will continue to actively take advantage of AWS services. Under consideration, among others, are plans for Containerizing with Amazon ECS (Elastic Container Service) and Amazon EKS (Elastic Kubernetes Service), and utilizing Amazon Connect in At Home call centers.
“The great appeal of AWS is its full range of services which leverage cutting-edge technologies to drastically reduce development burdens. While continuing to take advantage of these benefits, we plan to transform our organization, to be able to provide services novel to the world at even greater speed,” adds Yamamoto.
About At Home Co., Ltd.
- Established: December, 1967
- Capital: 100 million JPY
- Employees: 1,575 (as of end of February 2021)
- Business Description: Information service for real-estate companies, real-estate information service for consumers, real-estate business support services
Benefits of AWS
- Cost savings including licensing fees
- Ensured scalability to cover growth of users and data volume
- Reduction of development time for new applications to one tenth
- Reduction of operational workload
AWS Services Used
Amazon Elastic Compute Cloud (Amazon EC2) is a web service that provides secure, resizable compute capacity in the cloud. It is designed to make web-scale cloud computing easier for developers.
Amazon Aurora PostgreSQL
The PostgreSQL-compatible edition of Aurora delivers up to 3X the throughput of standard PostgreSQL running on the same hardware, enabling existing PostgreSQL applications and tools to run without requiring modification.
Amazon RDS for PostgreSQL
PostgreSQL has become the preferred open source relational database for many enterprise developers and start-ups, powering leading business and mobile applications. Amazon RDS makes it easy to set up, operate, and scale PostgreSQL deployments in the cloud.
AWS Schema Conversion Tool
The AWS Schema Conversion Tool makes heterogenous database migrations predictable by automatically converting the source database schema and a majority of the database code objects, including views, stored procedures, and functions, to a format compatible with the target database.
Companies of all sizes across all industries are transforming their businesses every day using AWS. Contact our experts and start your own AWS Cloud journey today.