Municipalities Upgrade Storage to Maximize Scalability

Unlike private companies, governments typically can’t make the case that tech investments are driving new revenue, says Brent Ellis, a principal analyst at Forrester. Combined with budget constraints, he says, this means that public-sector organizations tend to be less proactive with infrastructure upgrades.

“My experience is that people don’t upgrade storage until they are running out, or until something goes wrong,” Ellis says. “State and local governments are more sensitive to budget constraints, and they’re more likely to let a system stay in operation longer than it should, because they don’t have the funding to replace it.”

Mesa Deploys a A 10-Year Data Storage Solution

Mesa’s legacy storage system had high maintenance costs, and it was experiencing some performance degradation, which impacted core applications, including financial systems, permitting and public safety databases. The Pure Storage array provided significant improvements in speed, reliability and data reduction, which allowed Mesa to streamline operations and improve service delivery to constituents.

CDW Government led the design and deployment of the new Pure Storage environment, and city officials were able to manage the migration with zero downtime.

“CDW has been a tremendous partner,” says Scott Conn, the city of Mesa’s CIO. “They have been very responsive, and they’re willing to jump in and get their hands dirty with our team and work on the configurations. They have been absolutely blue-chip, excellent partners.”

Even before the move to Pure Storage, Mesa was running all-flash arrays. The upfront capital cost is not much more than that of spinning-disk storage, and Mesa IT leaders note that flash takes up less space, with lower operating costs for power and cooling. Among providers of flash arrays, Wilson says, Mesa leaders viewed the Pure Storage feature set as best in class.

One example: simplified synchronous replication between storage arrays, which allows the city to copy storage from one array to another in real time. “Other vendors offered some of those features, but they were difficult to implement,” Wilson says.

IT leaders were also drawn to the Pure Storage upgrade program, which gives the city new controllers for its storage environment every three years, extending the expected life of the array to 10 years. “We try to go to the well one time, do it right and build some runway and scalability into our thinking,” Conn says.

Achieving ‘Tenfold’ Speed With Flash Storage

When the COVID-19 pandemic hit and Tyler, Texas, sent city employees home to work remotely, officials asked: “What are we going to do?”

“I said, ‘Nothing,’” recalls CIO Benny Yazdanpanahi. “I told them to go home, log in and get to business.”

Fortuitously, the city had recently moved to virtual desktop infrastructure, which supports fast, secure connections to city resources, even for remote workers. To support the implementation, Tyler adopted IBM flash storage. Over time, the city has continued to support its VDI environment with a mix of all-flash arrays from IBM and Dell, while using less expensive spinning-disk storage for archival data, such as video.

“In the past, everything was spinning disk, and so there was latency,” Yazdanpanahi says. “It was reliable, but it was not as fast as we wanted. We needed the fastest storage. That’s the only way to do it for VDI. You cannot have latency. You have to be able to provide quick service. The flash storage was tenfold faster than what we had in place before with spinning disk.”

Storage Built for Exponential Growth

With the exception of a few straggler laptops, Tyler’s entire production environment has moved to flash-supported VDI, Yazdanpanahi says. Additionally, the city uses Microsoft SharePoint to store a number of its documents in the public cloud.

For each use case, Yazdanpanahi says, city officials analyze whether the benefits of faster flash storage justify the higher costs.

“We look to see what we are trying to do,” Yazdanpanahi says. “We ask our partners to show us the value. If flash is too expensive, I’m going to say, ‘No, this is not buying me anything.’ Video is one example; we don’t need to go to flash for that. It’s just too expensive, and it’s not cost-effective. I have to ask how often we access the data.”

By contrast, Yazdanpanahi says, speed is crucial when dispatchers send photos of suspects to police officers in the field, when emergency service vehicles connect to mapping software or when building inspectors pull up information that helps them with code enforcement.

DIVE DEEPER: On-demand storage is a game changer for state and local government.

“You need to be able to push data to police officers immediately,” Yazdanpanahi says. “Because we’re sending a picture through VDI, it doesn’t have to hit the spinning disk, and it goes out to patrol cars instantaneously.”

The exponential growth of data, he says, requires local governments to constantly re-evaluate and optimize their storage environments to adapt to changes in the market as well as changes in their own needs. When Yazdanpanahi started working with Tyler in 2001, he notes, the city stored about 700 gigabytes of data, which would have fit on about 700 storage CDs. Today, he says, Tyler stores more than 1 petabyte of data, which would require 1.5 million CDs — a stack nearly as high as the world’s tallest building.

“Now AI is coming, and that’s another beast in itself,” he says. “The hunger for storage is growing. It’s not going to go away.”