During a recent Qumulo webinar hosted by Alessandro Dellavedova, Senior Systems Engineer at Qumulo and Kim Buts, IT Systems Team Lead at Imelda Hospital, they examined a changing landscape for healthcare providers—particularly hospitals. They’ve seen exponential growth of data, particularly unstructured data, created daily from modern diagnostic and imaging equipment. Legacy storage solutions, with dual-controller architectures, have struggled to scale with growing data demands, resolve capacity issues, and meet performance needs.
New picture archiving and communication systems (PACS), digital pathology, and DNA sequencing are exponentially increasing hospitals’ file sizes. They’re observing a 1000x increase with 2D to HD images as doctors opt for improved, faster diagnostic equipment that helps them deliver better patient care. Digital pathology equipment generates upwards of 150 data gigabytes per case and DNA sequencing creates nearly 200 gigabytes per equipment piece. In the past, that meant adding more drives to account for storage pressures and reduced capacity, which led to increased data silos, critical performance issues, and an inability to scale indefinitely.
As Imelda Hospital upgraded its AGFA PACS last year, it was a slow, costly, multi-server system running locally in two data rooms with external storage for disaster recovery. It demanded significant resources and time. As Buts explained, “Thinking about our old setup gives me shivers.” Several parts had to work right or there were issues and they needed to carefully consider and plan under their previous capacity pressures so doctors could access patient images. The cusp for stored images was so high that raising capacity was technologically and financially challenging.
“Retrieving images from the external VNA was extremely slow … and retrieving lots of images took time—close to 45 minutes.” – Kim Buts, IT Systems Team Lead, Imelda Hospital
Growing storage requirements for Imelda Hospital
The priorities for Imelda as they assessed storage solutions to support their upgraded PACS were: easy-to-expand capacity, simplified setup, implementation, and maintenance, immediate failover, better performance and support, and finally, having snapshots for Cryptolockers.
With their small team required to manage the hospital’s complex and growing data files, Imelda wanted a solution with no file size limits, clear data and storage performance visibility, and an increase in the application’s efficiencies and the user experience. Any file storage downtime can be costly to the hospital, providers, and most importantly the patients.
With careful examination and testing, Qumulo and Fujitsu answered these critical needs and more. During the webinar Buts shared, “The system is so fast that users are not complaining anymore about latency or slow images. When users are happy, system admins are happy too!”
Qumulo fulfills Imelda’s evolving data demands with fast, effective performance
Imelda’s three Qumulo clusters—two in their main server rooms and for disaster recovery, plus another that is disconnected for image replication or archiving—were set up in less than four hours. The only maintenance is a monthly software system update that takes approximately 15 minutes. With Qumulo and Fujitsu, there are no performance disruptions, storage scales indefinitely, and IT always sees PACS server activity in near real-time to help with file optimization, planning, and troubleshooting.
Imelda dropped its expensive tiering by separating the Qumulo cluster for replicating images, which helps increase application speeds, performance, and supports cost savings. On their legacy solution, there was 100 milliseconds of latency before PACS images were processedimages processed and now latency is never more than 13 milliseconds. Doctors aren’t worried about images stalling, or worse, getting stuck. The critical information needed to treat patients isn’t stalled in data neverland and providers know Qumulo will effectively and quickly retrieve their files as they focus on delivering excellent care.
“We’ve gained time to focus on urgent, higher priorities because everything is extremely fast. Even zip archive users don’t see latency when images come from the zip archive or hot cache; it’s all the same.”
Imelda has also found support from Qumulo to be “mind blowing”, with significantly faster responses than other vendors they’ve worked with. Staff address their cases through Slack in five minutes or less and proactively identify and notify Imelda if they notice concerning issues such as a broken disk.
The next steps to evolve their Qumulo engagement include a shift of SMB shared data to a new cluster and migrating more images from DNA sequencing and digital pathology. Cloud data storage is potentially on the horizon, too.
Listen to the recent webinar or watch a recent video to learn more details of Imelda’s experience, the advantages of software-defined storage (SDS) from Qumulo to solve space and performance limitations that many healthcare providers struggle with, and helpful industry insights from Dellavedova and Buts.
