The Company: GAI Consultants
Migrating your entire Enterprise stack to the cloud can be complex and yield significant results. Mr. Michael Owens, GISP at GAI Consultants shares his experiences, lessons learned and advice on what led his organization to move their GIS to the cloud, the ins and outs of migration, and the benefits realized since making the move.
GAI Consultants delivers award-winning engineering, planning, and environmental expertise to energy, transportation, development, government, and industrial clients worldwide. With over 14 years of experience, Mike Owens, GIS Manager and Project Manager, specializes in managing GIS professional services and ad-hoc GIS support projects. He is responsible for leading a group of 12 GIS specialists, overseeing the technical work quality, improving company wide standards and workflows, and advancing GIS technology within the company and for its key energy customers.
The Problem: Performance & Expertise
Prior to cloud implementation, GAI faced issues with performance. With a large GIS staff across multiple physical office locations, trying to ensure optimal access to the GIS server, ArcGIS Portal server, and SQL database server proved challenging. All servers reside at their headquarters in Pittsburgh, PA and trying to gain access to those servers through a series of VPN tunnels can lead to instability and inconsistent performance. So for the employees not physically located at the central Pittsburgh office, they were struggling on a daily basis.
Another challenge centers around not having the IT resources available to become in-house IT GIS experts. IT resources that are formally trained in Esri’s best practices, the architecture, design, installation, configuration, maintenance and upkeep of the GIS systems. As Mr. Owens explains, “you really need to know what you’re doing to be able to be effective. Even with training, you have staff turnover and have to deal with knowledge transfer and retraining as well as keeping up with the technology as it moves forward”.
GIS technical leaders can address some of these challenges but in order to do so, IT support and access is needed. Each company’s policies and the IT department structure is different so this can add additional complexity and challenges. Additionally, IT resources are company wide and have customer facing initiatives that often and should take priority.
“Also on-premise servers are pretty expensive, not only to buy, but just to maintain and upgrade. We’d have to replace them. We often found ourselves a little bit further behind in the hardware technology and that ultimately caused issues with some of the Esri software we were trying to update. We’d run into issues where the hardware couldn’t really keep up.”.
The Solution: Migration to the Cloud
For Mr. Owens and the GIS leaders at GAI, once they had the idea they might want to move off-premise, they put together a cost benefit analysis to determine whether it made sense financially to move to the cloud. They started to quantify as much information as they could.
How many hours are lost due to inefficiencies like poor interoffice network performance? How much potential revenue lost did that account for? How much would it cost to hire a full-time dedicated IT staff member to manage the stack on premise? Could it be done with a part-time resource? Is there someone we could leverage internally? If we have a dedicated GIS IT resource, what additional costs are there to maintain and replace servers as they get outdated and can’t handle the needs of GIS. How much time and resources are spent running backups? “If you’re not doing backups in the cloud, it can be very costly. Just even having the resources available to perform a disaster recovery late at night or something like that is important and really pushed us hard in the direction of the cloud.”
Want to calculate your organization’s ROI?
“Our biggest concern moving into the cloud was data security. Once you’ve moved your data off-premise, make sure it’s secure. Some other concerns included confidence in the uptime of the system, integration with our company standard policies, like managing what a user can and can’t do when they’re logged into a machine, compatibility with other software and our other servers, our accounting systems and project management systems, Skype, and Microsoft Teams. We needed to make sure those would all work post migration. We also want to be able to maintain a certain level of control and access to our system.”
There’s so much that goes into a successful, performant GIS and with all the considerations, they answered key questions. Should we do this on-premise? Should we hire someone? Should we move this to the cloud? Should we have someone else help us manage it? Mr. Owens and the GIS team at GAI interviewed several cloud-based GIS IT partners and with the support of their ROI calculations, decided to move to the cloud.
The process for GAI was simpler than originally thought. “We filled out an assessment, a questionnaire that outlined our goals and our challenges. Our requirements provided information about our existing environment – what our servers look like, how much data we had on them, what kind of software we run as well as what our users need, what types of users we have and how did they use GIS on a daily basis. It was a deep dive into how you envision your GIS to look. This got us thinking about the nuts and bolts that go into an implementation of this magnitude”.
GAI’s GIS department had the support of their IT group and the understanding and collaboration of their CIO. Having buy in at the C-level made the transition even easier. It also brought GIS to the forefront of the company overall. With this attention also came added pressure to ensure the implementation team was doing their due diligence and being extremely thorough, especially in the early phases of the migration process.
After completing the cloud assessment, key to evaluating scope and level of effort to complete a cloud migration, GAI’s GIS team had working sessions with ROK Technologies, their managed services cloud provider, to dissect GAI’s assessment. “We asked each other a bunch of questions and started the process of putting together a work plan to get to a cost. We reviewed ROK’s architecture and design implementation plans for the system including how the servers are connected and how users are going to connect to the system. From there ROK took care of the rest. They set everything up; migrated our databases from the old servers to the new; built out our virtual machines. Of course, we were still afforded all the opportunities to collaborate with them along the way and test different components while in development.”
GAI also leveraged their internal IT resources to work through any challenges specific to their organizational needs. The hardest part of the migration process was some of the technical challenges like firewall and network issues. He advises on how important it is to work closely with internal teams and external partners to “figure out how to get your networks connected to an outside resource, like a Citrix or an AWS solution”.
Mr. Owens shares that there is a lot that goes into planning and executing a cloud migration and you can expect some hiccups along the way, but for GAI, the easiest part was choosing a partner for their virtual environment. ROK Technologies set up demonstrations of the different options available for their needs along with pricing in an effort to make the most informed decision. Ultimately GAI’s GIS division selected AWS once they saw what it could do. “It was so easy for us to just say this is exactly what we want and what we need. We just made the decision and ran with it”.
The Benefit: Didn't Skip a Beat!
The distributed GIS workforce had struggled for so long in the on-premise environment that on day one in the cloud, they recognized improvements. Now everyone, whether situated at headquarters or remotely, had the same level of performance. This boosted utilization rates that we can tie back to our revenue. The GIS division also saw a decrease in the reliance of their IT department. GAI’s IT no longer needs to respond to on-premise server issues, schedule patches, conduct upgrades or submit and manage Esri IT help desk tickets. Instead ROK Technologies handles the GIS IT work allowing GAI’s company wide IT department to focus on their own help desk log, data and network security improvements, software rollouts and other higher priority initiatives. Mr. Owen shares, “we have ROK support 24/7. Anytime we need anything, we get a fast response and we have confidence that the solutions are rooted in knowledge of Esri’s best practices”.
A more unintended success story is how the GIS group and GIS users across GAI haven’t really missed a beat since the pandemic forced everybody into working remotely. By logging into a virtual machine and having your desktop in a virtualized environment, all the servers and data needed is available at their fingertips. GIS staff members can be productive from anywhere as long as there is internet.
The GIS team plans to replace their big, heavy, expensive GIS laptops with a thin client this year because they are in the cloud. “Now that all the horsepower is in the cloud, we don’t really need a $500 graphics card to do 3D GIS work. So getting everything into the cloud is going to save you money. I think your IT departments are really going to love you just for that.” All the high end processing takes place in the cloud where CPUs are cheap, RAM is cheap, and storage is cheap. With server side GPUs, a GIS workforce can work off a Chromebook or Mac, or a really thin client and ArcGIS Pro will work like a dream.
When GAI brought on a new hire that needed access to GIS software pre-cloud, the process took time and resources – assigning a physical laptop and installing the software needed. Now post-cloud, “we just open up the Azure active directory in the cloud and add their login name to the AppStream user pool and they are up and running.”
In Conclusion: It's Important to Do your Homework
Have a good, solid grasp on what it is you’re looking to do and why you need to do it. Calculate the return on the investment for your organization. Finally talk to people – your GIS team, IT resources, and importantly cloud, GIS and IT experts to help answer some of the big questions:
- Is a move going to free up your GIS team to do more for your customers or citizens?
- How much does it cost to refresh on-premise servers every year?
- Do you need to move all of your servers to the cloud?
- Should you take your current environment in your server room and just pick that whole thing up and move it to the cloud? Is there a reason to migrate or lift and shift that exact server to a cloud environment?
- How can we improve, enhance the existing architecture moving forward in the cloud?
- Do you want ArcGIS server and Portal on the same machine? You could put map cache on Amazon S3, and it would be much less expensive compared to storing it on your local storage.
It is vital to have both GIS and IT knowledge, training and know-how to architect a solid GIS solution in the cloud. Finding a good IT person is tough. What is even harder is finding a good technical GIS person that can do it all.
Jason Harris, CTO at ROK Technologies shares:
“What we want to do at ROK Technologies is to become an extension of your GIS department. We want to be your trusted advisor to free your IT guys up as well as free up your GIS staff up to focus on GIS products and services. We also want you to be in a strong position to take advantage of your Esri investment and the different extensions the enterprise rolls out; and to help you utilize the licensing that you’re paying for.”
Note: This content was featured in our webinar, 2020 Esri IMGIS Panel Discussion: The Move from On-Premise to the Cloud, featuring Mike Owens along with GIS leaders from MassDot and Connecticut Water.