Software delivers perfect prescription: healthcare services firm strengthens its offsite data backup and recovery operations.
"We wanted to create a very robust data-protection strategy with multiple redundancies," says Chris Baldwin, vice president of corporate MIS for Northeast Health. "We're talking about protecting applications, data and images that our physicians and nursing teams use every day to treat patients. We can't afford to take any risks that may affect the level of patient care."
The new approach was to create a system with local, remote and full disaster-recovery points. The objectives were to reduce recovery time, minimize data loss and system downtime, and to fully automate as many processes as possible. Whatever approach was selected needed to protect Northeast Health's two data centers and 35 remote sites, comprised of hospitals, nursing homes, home care, supportive living, rehabilitation and community service agencies.
Northeast Health was introduced to backup, archive and recovery specialist BridgeHead Software via Dell Computer's healthcare division and its system integration group, Dell Professional Services. BridgeHead's HT Backup software performs automated, unattended backup and replication of data, databases and applications.
HT Backup provides application-integrated support for Northeast Health's existing healthcare information system, MEDITECH, through its integrated serverless backup (ISB) and integrated disaster-recovery modules, which are designed for operation in a storage area network (SAN) environment. Other key parts of the system include 12 live and six standby Dell 2650 servers running Windows 2000, an EMC CX-700 storage area network and a Dell PV136T tape library.
DATA BACKUP AND REPLICATION
At the core of the new disaster-recovery strategy is fast data backup and data replication between the primary data center in Troy, N.Y., and the secondary site in Albany, N.Y., seven miles away. The sites are connected via redundant line-of-sight 100-MB and 150-MB microwave links across the Hudson River, as well as a 10-MB virtual private network connection. In the event of a failure of the primary circuit, the network will automatically fail over to the secondary and/or tertiary circuits.
BridgeHead's technology was the first solution Northeast Health found that could successfully support a wireless link. This was an important benefit, as the alternative of deploying Fibre Channel connectivity would have been prohibitively expensive.
Using the ISB module for MEDITECH, HT Backup creates a point-in-time copy of Northeast Health's 12 MEDITECH servers. The data copy is immediately compressed and transported across the microwave link to the remote center, where it is written out to staging disk. The copied data is now available to be brought back across the network in the case of point failure and the need for operational recovery.
The instant each data copy is received at the remote site, it is immediately restored to a go-live state on additional disk devices accessible by spare standby MEDITECH servers. This gives the hospital the ability to go live on consistent and functional data in an immediate response to a disaster at the primary location. Finally, the data replicas are also automatically backed up to tape using a tape library at the remote site and allow for removal of data for long-term storage.
BridgeHead's ISB is the key feature that lets MEDITECH users take advantage of SAN "snapshot" capabilities (also called clones, snap copies or flash copies) by creating static, point-in-time images of the entire MEDITECH environment.
BACKUP IMPACT IMPROVED
Within Northeast Health's system, HT Backup sits on a dedicated backup server and orchestrates the backup to create image-based backups of each of the SAN-based logical disks. It controls the backup schedule and communicates with MEDITECH's application when a backup is to start. MEDITECH then locks down, releases each of the SAN mirrors, and hands a list of those mirrors to HT Backup, which makes an inert, static image of each mirror to the tape library. Once done, HT Backup informs the MEDITECH application and the mirrors are re-synched back to the HCIS.
In addition to business continuance, the new system for protecting MEDITECH data and availability has improved the operational impact of backup on the organization. Previously, clinical staff, pharmacists and physicians had complained about a slowing in the performance of the healthcare information system during the five hours or so when the manual backup process was performed. With the ISB technology, Northeast Health is now able to minimize data loss by running frequent, automated backups in waves of just 35 minutes, all without users noticing any impact on the day-today performance of the system.
The entire new three-step disaster-recovery process of replication, restore and backup now takes just four hours, less than the previous manual single-step backup.
Northeast Health was able to secure its data at 70% of the cost spent the last time it upgraded its hardware platform seven years ago. By replacing the traditional direct-attached storage system with a SAN and purchasing BridgeHead's HT Backup, it now has a rigorous disaster-recovery system with many more features than expected.
"The solution helped us solve two problems at once," Baldwin offers. "We get to automate our daily backups, and we get the disaster-recovery solution we've been looking for. That kind of integration is by far the biggest benefit, and something we couldn't find anywhere else."
Northeast Health plans to expand its use of BridgeHead's technology with the rollout of HT Backup to protect the picture archiving and communications system (PACS).
For more information from Bridgehead Software: rsleads.com/605cn-252
RELATED ARTICLE: Make backups continuous.
by Chuck Shavit
Most backup technologies employed today are not remote, and they are not continuous. Until recently, backup has been done only at predefined times, mostly automated to the middle of the night. As such, the dreaded "backup window" problem has existed--where changes made after the last backup are at risk of loss. In addition, the automated process can fail, and go undetected.
Continuous data protection (CDP) solutions for disk-to-disk (D2D) exist today that handle backup as the data changes. Such CDP solutions deliver "any-moment-in-time" protection, and also eliminate the backup window.
In many scenarios, the benefits of consolidated CDP backup are significant for:
* local or remote offices or sites where the data is critical for continued operations;
* remote offices where data is important to other locations, or for compliance;
* offices with more than five users;
* offices with no IT staff on-site;
* offices that require a high level of daily data availability for operations to continue; and
* users who often call IT for deleted or past versions of files.
Not all CDP solutions are capable of remote backup, nor are they continuous. Remote CDP turns out to be a major challenge if the distance between the source and the backup location is non-trivial. For example, an enterprise might have data in several cities in the United States, in Europe and in China. If all these locations need to back up to the corporate headquarters in the United States, and if the backup is to be continuous, bandwidth issues might render the process impossible.
Traditionally, CDP products come in two flavors: CDP of unstructured files, such as spreadsheets, CAD drawings or legal documents; and CDP of structured files, mostly databases. Most CDP products today are optimized for structured data, and will thus transfer several times the data size, for even the smallest change. Such approaches solve one issue--nightly backup with a backup window--and introduce others- network slowness, over-utilization and server performance degradation.
Conversely, if a CDP product that is optimized for unstructured files backup is used for databases, the database would be backed up only when the file is closed-thus it is not continuous.
When evaluating CDP solutions, identify those "generic" CDP technologies that may be fine for local use but not be appropriate for remote offices. Here is a checklist of backup/recovery features that provide the best level of protection at the remote office:
Failover and fail-back capabilities. With fail-back capabilities, any changes made during the failover period (while users access their data on the failover system) are instantly reported to the original server when it comes back on line.
Overcoming network concerns. Some CDP products offer byte-level differencing, compression and encryption techniques that can eliminate 95% of bandwidth used in the remote backup consolidation to a data center. While this is helpful for local CDP environments, it is paramount for cost-effective remote office backup consolidation.
Proper support for specific file types. Hot database technology will not support CDP for file backup over distance, and file backup will not work at all for hot databases. If your remote site needs both, you should factor this accordingly into product evaluations.
True CDP vs. preset restore points. Most requests to restore lost or corrupt data occur within hours-far sooner than the last backup. Many CDP products are scheduled snapshots. They do not capture each version or each transaction. They cannot access any past version, or roll back a database to "any moment in time."
Chuck Shavit is CEO of Availl, Andover, Mass.
For more information: rsleads.com/605cn-251
|Printer friendly Cite/link Email Feedback|
|Date:||May 1, 2006|
|Previous Article:||Storage insurance: Dan Simpson, CIO at Physicians Mutual, engineered his firm's launch of a major business transformation initiative, including the...|
|Next Article:||E-mail challenges storage environments: the regulatory need to archive and manage records calls for advanced capabilities.|