Database Case Study
Disaster Recovery for the Emergency Department Application at St. Vincent’s University Hospital
St Vincent’s Healthcare Group, (incorporating St Vincent’s University Hospital, St Vincent’s Private Hospital and St Michael’s Hospital) provides acute general medical care serving the South East region of Dublin and surrounding areas.
St. Vincent's University Hospital is a major academic teaching hospital. They provide a front line emergency service and national/regional medical care at inpatient and outpatient levels with an excess of forty medical and surgical specialties.
DNM Technology is an independent IT infrastructure company who specialise in databases, virtualisation and intelligent data storage & management. Working with our customers we can provide database and application architecture design, implementation and support services through our Database Assurance Services.
“We see it as one of our core objectives to build and maintain a stable platform for all of our systems.
This constitutes a key component of disaster and continuity planning, in particular for our core clinical systems. To this end we periodically perform health-checks and reviews of well established systems such as our Emergency Department (ED) system.” Says Mr. Dermot Cullinan, Director of ICT.
“As part of St. Vincent’s ICT Department disaster recovery strategy, we were reviewing our plans in place for our core clinical applications. We wanted to ensure that all core clinical systems were designed to offer high availability and had a full disaster recovery plan documented. When we reviewed our system we identified limitations with the version of Sybase we were running, it was not cluster aware and there was no upgrade path as our application only supported the version we were on. We then turned to DNM to plug the technical gap we had with our Sybase skills” says Neal Mullen, Deputy Director of ICT.
“The hospital has recently been awarded JCI Accreditation. During the audit of SVUH the JCI team focused on a number of our clinical systems including the Emergency Department ICT System. JCI were complementary of how we managed our disaster recovery planning and support of our ED System. Working with DNM on our ED Project gave us greater confidence in achieving effective support and disaster planning. DNMs service has been flawless over the past 18 months they have been supporting our ED System” adds Mullen.
SVUH approached DNM Technology to carry out an Applications Health check. The health check revealed problems with our Disaster Recovery provisions that could lead to outages of over an hour with the existing solution which could leave the department without their application should the primary hardware failed.
“The Emergency Department system is a comprehensive solution which tracks our patient’s journey from registration through triage to discharge, recording key evaluations. Our Emergency Department moved to a purpose built location in our new clinical services building. Because of the additional size of the department the ED System became even more vital, particularly its patient tracking component”. Says Dr. John Ryan, Senior Consultant in Emergency Medicine.
“We needed to ensure that we were offering ED the best architecture possible for their system. We wanted to ensure minimum data loss (15 minutes data) in the event of a problem and were able to restore the ED System within five minutes regardless of the hardware or software issue.” Says Neal Mullen
SVUH subsequently engaged DNM to work with our Operations Team to architect an improved disaster recovery solution for the ED Departmental System, one that would meet both their Recovery Time Objective (RTO) and Recovery Point Objectives (RPO).
Working with the stakeholders, SVUH Operations Team and DNM architected a solution which included elements of Database, Hardware and SAN storage technology; this solution would provide all the recovery time objectives and recovery point objectives desired.
The solution included:
· Sybase Full Backup which occurs once a night along with seven day backups maintained on local and redundant server
· Sybase Incremental Backup which occurs every 15 minutes along with seven day backups maintained on local and redundant server
· Standby Server which is a blade identical to the primary server exclusively maintained for failover/replacement
· SAN Storage for all disks - No local disk storage meaning the standby server can replace the primary in case of server failure
· Data Volume Snapshots which is taken 4 times a day
