We were able to build a very scalable back end to support millions of images using Amazon EBS sc1 volumes, and the number of images is growing daily. Using AWS with EBS volumes, we know we can scale to meet future growth.
Pål Kristensen Head of System Operations and DevOps

Geodata is a leading provider of geographical information systems (GIS) software and technical and professional services across Norway. The company specializes in map-data analytics and visualization solutions based on Esri software. It helps public and private sector customers make better decisions based on data and perform other vital tasks every day.

  • Needed to cost-effectively store millions of aerial GIS images for a new website based on Esri map visualization software. The company also needed to support a Windows-based file system to process more than 120 terabytes of raw aerial image data for the website.
  • Needed to store more than 48 terabytes of processed and compressed aerial images on a file system natively accessible by Windows Server.
  • Wanted to ensure fast performance for the site’s map-data processing and also deliver high availability and reliability for the site’s end users. “Many individuals and organizations rely on the aerial images served by the website, so we need to consistently deliver a great user experience,” says Pål Kristensen, head of system operations and DevOps at Geodata.
  • Needed the ability to scale the website quickly and easily as the volume of data and number of users continued to grow after the site’s launch.
  • The new aerial mapping site, norgeibilder.no, provides millions of high-resolution GIS images. The processing engine for image storage and caching includes a file-storage system based on Amazon Elastic Compute Cloud (Amazon EC2) and Amazon Elastic Block Store (Amazon EBS) Cold HDD (sc1) volume types. The Amazon EBS sc1 volumes are designed for large sequential I/O, high-throughput data workloads.
  • Map data is sent to a web server fleet from Windows Server over a Windows Common Internet File System (CIFS) share. The file-storage system includes 48 terabytes of map tiles stored on four-terabyte Amazon EBS sc1 volumes, which are attached to an Amazon EC2 Windows instance.
  • Geodata configured a write cache running on Amazon EBS General Purpose SSD (gp2) volumes, using Windows Storage Spaces to buffer data write peaks. Additionally, the company archives 125 terabytes of raw aerial images on Amazon Simple Storage Service (Amazon S3).
  • Reliable storage. With its file-storage system based on Amazon EC2 and Amazon EBS sc1, Geodata can quickly and reliably store millions of map images from throughout Norway. The aggregate system can support read capacity of 500MB per second. “The Amazon EBS sc1 volumes give us the performance we need to support the site’s workload,” Kristensen says.
  • Costs lowered by 70 percent. By using Amazon EBS sc1 volumes, Geodata has lowered its storage costs by more than 70 percent, and their overall AWS monthly costs by more than 40 percent. “We pay considerably less for the Amazon EBS sc1 volumes than we were paying before,” Kristensen says. “Overall, supporting a system like this on-premises would be much more expensive and complex."
  • Scales to support growth. Geodata can easily scale its file-storage system and website to support growth in number of map images and number of site visitors. “We were able to build a very scalable back end to support millions of images using Amazon EBS sc1 volumes, and the number of images is growing daily,” says Kristensen. “Using AWS with EBS volumes, we know we can scale to meet future growth.”

To learn more about how AWS can help you manage your website, visit our AWS Website/Web App details page