Managed Data Support Case Study
Now we no longer have to worry about the risk of downtime during our busy conference registration period, or at any time. Our data centers are completely fail safe, without worry.
JACK SPENCER, CIO AT ASHP
Overcoming Shaky System Availability
To connect pharmacists and introduce new technologies, ASHP hosts multiple meetings and conferences throughout the year, including the ASHP Midyear Clinical Meeting & Exhibition — the largest gathering of pharmacists in the world.
At the height of the busiest registration period, ASHP’s old IBM storage area network (SAN) began to fail. All vital registration data was at risk. The team had to restart the system’s controllers three or four times a night just to keep registration open.
The association needed to collect critical registration data, accounting for more than 30 percent of the organization’s annual income — and risked losing millions.
“Just as we began to take our highest volume of registrations, our IBM SAN was getting pretty shaky,” said Jack Spencer, CIO at ASHP. “When we contacted IBM, they decided the controllers were at the end of their life. We didn’t have time to replace them at that time, we just wanted to stay alive. But we knew that once our critical registration time was out of the way, we needed a new solution for both reliable storage and failover for guaranteed availability.”
Preparing Critical Pharmacist Data For A National Emergency
Beyond ASHP’s system availability issues, its consumer website, www.safemedication.com was also at risk. Consumers rely on the site for critical drug information including medication tips and drug shortage information.
Because ASHP hosts the largest pharmacist database, the U.S. government also depends on it to distribute critical pharmaceuticals during national emergencies. For instance, the Food and Drug Administration could request pharmacist information to help distribute critical treatments – such as the anthrax treatment Cipro during the 2001 anthrax scare.
“For national safety reasons, our system must be available at all times,” said Spencer. “Our organization has a responsibility to the public’s safety and the effectiveness of our members. Our systems simply can’t go down.”
If a system outage occurred, ASHP needed a highly reliable, completely redundant storage solution so it could fail over to an identical solution.
Redundant and Resilient Storage Infrastructure
ASHP had previously worked with Helixstorm to deliver a virtualized storage implementation. This time, the organization needed a custom, fully redundant solution to remedy its downtime risk — with zero disruption to operations.
“Just as we made it through our busy conference registration period, our UPS system was starting to die, along with the challenges we were having with our IBM SAN,” Spencer said. “We needed to renovate our storage infrastructure, but we had to do it without any business impact or loss in our production uptime.”
Helixstorm architected an innovative storage virtualization and data protection solution to keep all information available and avoid downtime.
Designing A Fully Redundant Storage Solution
The ASHP datacenter occupies two floors in its Bethesda, Maryland, office building, so Helixstorm designed the infrastructure for synchronous data replication between floors.
During implementation, one floor would be active while the other was overhauled. The systems would then shift over to the identical solution on the other floor while Helixstorm deployed the data center on the initial floor.
“We structured the solution to dynamically move workloads between ASHP’s data centers on the second and third floor without any risk of downtime,” said Aaron Schneider, co-founder and principal engineer at Helixstorm. “Overall our goal was to increase data resiliency by physically separating resources between floors.”
To achieve this, Helixstorm implemented two virtualized storage systems. Each solution is replicated on both floors, giving ASHP high reliability through synchronous data replication.
Helixstorm successfully implemented the solution without any downtime or impact to ASHP’s production environment. As planned, Helixstorm dynamically moved workloads between data centers so the storage infrastructure and new UPS systems could be implemented without interrupting the business.
With a synchronous data replication solution that leverages ASHP’s virtual infrastructure, the organization has a true failover solution that assures business continuity and high data availability.
Preparing For Full-Scale Disaster Recovery
Next on ASHP’s agenda is a comprehensive disaster recovery strategy. Helixstorm will deploy asynchronous replication to an offsite data center, ensuring all data is available during a physical disaster or outage at ASHP’s headquarters.