In addition, PostFinance is taking advantage of the solution’s scalability and cost advantages to address a range of other data management needs.
The challenge of managing big data
Building on a storied 100-year history, PostFinance provides financial services to three million personal and business customers. One of its top goals is creating a customer experience across the digital and physical worlds. The company uses Hadoop and Apache Spark as the foundation for its big data analytics and found that traditional block and file storage could not provide the cost-effective scalability needed for storing this data.
“We decided we needed an object storage solution,” said Andre Kocher, senior systems engineer, PostFinance. “Starting with a list of major storage vendors, we conducted a thorough selection process and ultimately chose Cloudian.“
Cloudian’s limitless scalability, compelling TCO benefits and full S3 compatibility were differentiators.
Working with Infoniqa SQL AG, a Swiss integrator, PostFinance deployed Cloudian’s HyperStore object storage appliances in its two data centres, along with Cloudian HyperFile to support file services from the HyperStore platform. Data protection is provided across the data centres through Cloudian’s integrated policy-based data replication.
“The installation was very smooth, and, after testing the Cloudian solution with various applications, we put it into production a couple months ago,” said Kocher.
Modern storage foundation for multiple use cases
PostFinance’s priority was integrating HyperStore into the big data platform and moving data from Hadoop distributed file system to the Cloudian object storage. The company quickly began seeing the benefits.
“Cloudian’s integrated data protection means we don’t need a separate backup process and backup software licenses for the data,” Kocher reported. “This will save us money on our backup infrastructure over time.“
In addition to the big data use case, PostFinance is using Cloudian as the storage foundation for a 100 million file document repository and as an important part of the archive solution.
“We wanted to move away from a traditional file-based storage system to a more modern, flexible architecture, and Cloudian is an ideal way to do so,” said Kocher.
PostFinance is also exploring other use cases, including migrating log data (Splunk) from traditional block storage to Cloudian, employing HyperStore as the second-tier Splunk data platform. In addition, the company is considering a potential SaaS offering for externally hosted partner applications, providing shared S3-compatible storage from the Cloudian system.
“We have already run proof-of-concept tests with other S3-compatible applications and are very pleased with the results,” Kocher noted. “This year, we expect to add more applications and have up to 1PB of data under Cloudian management, making it a key enabler of our business moving forward.“