A Leader in Healthcare Staffing Improves Data Protection | Tintri

0 0

Cross Country Healthcare Improves Data Protection and Disaster Recovery with Move to Tintri Storage


Cross Country Healthcare Improves Data Protection and Disaster Recovery with Move to Tintri Storage
Cross Country Healthcare Improves Data Protection and Disaster Recovery with Move to Tintri Storage

CCH Reduces Recovery Point Lag Time from Over 2 Weeks to Less than a Day Using Tintri Snapshots and Replication

Cross Country Healthcare

Cross Country Healthcare (CCH) is a national leader in primary healthcare staffing and workforce solutions. Since its founding in 1975, it has placed thousands of highly qualified nurses and physicians, allied health, advanced practice, clinical research, and case management professionals into thousands of healthcare facilities across the United States and the Caribbean.

IT Challenges

Cross Country Healthcare had been relying on a mix of aging NetApp, EMC VNX, and XtremIO all-flash arrays before moving to Tintri. “We were using those storage systems for a lot of different workloads, including block-level storage, backup to disk, as a CIFS target, and then as a VM target,” explained Jason Bourque, vice president of infrastructure at Cross Country Healthcare. “Unfortunately, the NetApp systems didn't scale very well and we were always struggling to keep up with the virtual infrastructure demands. Ongoing management was also a challenge because we didn’t have any in-house VNX experience. The XtremIO systems provided much higher performance than the NetApp arrays, but they were far too expensive and difficult to manage.”

Bourque and his team evaluated several different infrastructure options, including legacy storage solutions, all-flash platforms, and Tintri’s hybrid flash storage. “I must admit, I was a bit biased going into the evaluation since I had used Tintri at a previous company,” Bourque said. “I already knew what Tintri could do—I just had to convince my team that it was the right choice for us going forward. After running a proof of concept on some of our real-life use cases, my team agreed that Tintri could provide all of the performance, flexibility, and stability we were looking for. We could solve all our existing issues and position ourselves for the future at the same time.”


CCH purchased its first Tintri hybrid flash systems in 2015. All production workloads were moved over to the new Tintri storage systems, enabling them to consolidate their IT infrastructure from five data centers to one. “We now have more than 400 different systems running on the Tintri systems, including our largest SQL databases and front office applications,” noted Bourque. “All of the disparate workloads are running harmoniously together on the same storage system, thanks to Tintri.”

CCH bought an additional Tintri system in March of 2017. “One of our larger business applications was still running on a NetApp storage array,” said Bourque. “That environment took up several terabytes of capacity, so we purchased an additional Tintri array for that application and a few other mounted systems. Now all of our production workloads are running on Tintri.”


Simpler Management

Bourque and his team were very impressed with Tintri’s management efficiency. “We used to have to hire dedicated storage engineers and cross-train everyone to provision LUNs and volumes on the NetApp and EMC systems,” Bourque reported. “With Tintri, our storage environment can be easily supported by our vSphere engineers, since they don't have to know anything about storage to operate Tintri.”

Higher Performance at a Lower Price

“The Tintri systems can easily handle the high I/O needs of our most demanding apps,” noted Bourque. “Even with our mixed workloads, we aren’t even close to exceeding the performance limits of Tintri—at a much lower cost than the all-flash alternatives. In addition to the 2x reduction in CapEx, there are several usability features that Tintri has that are not available with the XtremIO platform. The VM integration alone greatly simplifies the management of our growing infrastructure.”

Supporting a New Disaster Recovery Initiative

After migrating all of their production workloads over to Tintri, Cross Country Healthcare embarked on a data protection and disaster recovery improvement project in 2017. All of CCH’s production workloads are running in an Equinix hosted co-location facility in Miami, Florida, and the company’s privately owned datacenter in Atlanta, Georgia is being used for disaster recovery.

“We built our DR environment using VMware and Cisco blade servers,” noted Bourque. “We originally went with traditional software-based replication, using Veeam with ExaGrid as the underlying hardware, as opposed to storage-based replication. But since we have over 50 systems with very large data sets, there was a tremendous amount of lag in the system due to all of the changed data that had to be copied to the other site.”

The large, software-based backup jobs were also consuming far too much bandwidth, according to Bourque. “Our 100MB link to our DR facility was pretty much maxed out full-time,” he explained. “As a result, we had to ‘rate limit’ our replication traffic because we didn’t want it to impact our business traffic during the day. Due to the bandwidth limitations, we were usually two weeks behind in sending our backup data to the DR facility. Obviously, that’s not an ideal situation for security purposes.”

In order to improve DR, CCH moved one of its existing Tintri VMstore systems over to its backup center in Atlanta and purchased a new Tintri system for its primary data center in Miami. “I used Tintri replication at another company, so I already knew that we could improve on our RPOs if we put an additional system at our DR facility,” said Bourque. “By moving to Tintri, we were able to shorten our RPOs from two weeks to less than 24 hours. This removed a major business risk if we ever had to recover data in a critical situation.

Retention Schedule and Bandwidth Savings

“We are protecting all of our data with Tintri snapshots,” Bourque reported. “We keep local copies of all workloads and then replicate our production-critical systems nightly to our DR site Atlanta. With Tintri replication, we can complete all of the snapshotting within just a few hours.”

Bourque knew he could either increase bandwidth to fix the DR lag problem—or invest in Tintri storage. “Bandwidth is a lot more expensive than storage in the long run,” noted Bourque. “Storage block-based replication is much more efficient than software-based replication, where you are moving full data sets across the network. We’re saving a lot of money by using Tintri replication. On network bandwidth alone, it's a 2X savings. But the OpEx and manpower efficiencies we gained by not having to manage bandwidth every day are very significant. Choosing Tintri was a very simple and smart decision for our organization.”

Eliminating the Need for Multiple Management Tools

“The ability to easily clone and recover systems using the Tintri interface has enabled us to eliminate a lot of other management tools, including VMware’s SRM for disaster recovery,” noted Bourque. “The Tintri interface plugs right into VMware, so I just need someone who understands those two consoles to run our DR strategy. With IT infrastructure, I believe that ‘less is always more’. The fewer systems I have to deal with, the easier it is to support the environment.”

Future Plans

“When I first started working here two years ago, the infrastructure wasn't very reliable,” Bourque recalled. “I spent the first year stabilizing the environment and making it simpler by consolidating and standardizing on Tintri. Last year, I focused on improving DR. This year, I’m focused on business continuity. By using Tintri storage, we can ensure that we’re an enterprise-ready organization going forward.”

When asked if implementing storage-based data protection and disaster recovery on Tintri gave him peace of mind, Bourque replied, “Generally speaking, I am at ease by having ALL of my systems on Tintri—not just DR, but also our production systems. I’ve used Tintri for long enough that I am very comfortable with the high availability, recoverability, and management capabilities within the systems. I am a big believer in Tintri—it delivers all of the performance we need, in a much simpler and more cost-effective solution. We plan to run everything on the Tintri platform from now on.”



Healthcare Staffing and Recruiting


Boca Raton, Florida

Virtualization environment

VMware vSphere

Previous storage platform: NetApp, EMC VNX and XtremIO

VM profile

MS SQL databases and all virtualized applications

Key challenges

• Legacy NetApp systems were difficult to manage and scale

• XtremIO arrays performed adequately, but the systems were very expensive and lacked features

• Wanted to improve data protection and disaster recovery for business-critical applications and data

Tintri solution

Tintri VMstore hybrid flash systems

Primary use case

All databases and virtualized workloads for production and disaster recovery

Business benefits

• Reduced recovery points from >2 weeks to < 24 hours

• Improved storage performance

• Simplified storage management

• Eliminated the need for multiple management tools