Using ArcGIS Monitor to RightSize THE CITY OF HANFORD’s ArcGIS Enterprise on AWS
Written by: Brock Saylor and Chris Dougherty
...On California Climate Change Disclosure Rules
Written by Leslie Wong and Hussein Sayani
BACK TO HOMEPAGE
Share on Social Media!
Leslie Wong
Senior Associate
lwong@langan.com
Hussein Sayani
Senior Project Manager
hsayani@langan.com
Brock Saylor
Senior GIS Manager
For the city, ArcGIS Enterprise became more responsive for users. For Langan’s managed services team, the project showed how much ongoing monitoring and trend analysis can influence subsequent cloud architecture decisions.
Conclusion
ArcGIS Enterprise environments are not static—workflows, content, and usage patterns shift over time. Designs that are appropriate at launch may need to be revisited once real usage data is available. Through its managed services partnership with the City of Hanford, Langan used ArcGIS Monitor to turn operational data into concrete architecture decisions, rightsizing the AWS environment around actual demand. Beyond troubleshooting, ArcGIS Monitor is an ongoing tool for tuning ArcGIS Enterprise as an organization’s use evolves.
The City of Hanford, CA, relies on ArcGIS Enterprise to support a broad range of local government operations, including community development, public works, parks, and public safety. Across the organization, GIS supports workflows such as housing and zoning, public notifications, tank inspection forms, water distribution, stormwater and sewer operations, parks management, fire services, and pre-incident planning.
To support these needs, Langan provides ArcGIS Enterprise-managed services in AWS for the City of Hanford. As an Esri Gold Business Partner with the ArcGIS Cloud Specialty, Langan helps clients design, operate, and optimize cloud-hosted ArcGIS Enterprise environments that fit how their organizations operate day-to-day. Throughout our partnership with the city, we have worked closely with Jason Waters, Deputy City Manager; Maddison Faria, Assistant Planner; and Eric Forcey, IT Manager. ArcGIS Monitor became a central part of this work, replacing assumptions about how the environment would be used and enabling observability for ArcGIS Enterprise.
Original Architecture
The city’s ArcGIS Enterprise environment was originally designed around the expectation that most GIS services would be published from enterprise geodatabases stored in PostgreSQL. Based on that assumption, the architecture included a base deployment of ArcGIS Enterprise, a federated ArcGIS Server site intended for SQL-published services, Amazon RDS for PostgreSQL, web adaptors, an Application Load Balancer, and Amazon CloudFront.
The design was reasonable based on the available project information and experience with comparable clients. The environment separated anticipated workloads and provided a dedicated path for services published from enterprise geodatabase data.
After nearly a year of operation, usage shifted. Instead of publishing services from SQL, staff members were creating and consuming hosted services, and hosted feature layers had become the primary workflow. This increased demand on the hosting server and ArcGIS Data Store, while the federated server designed for SQL-published services was underutilized.
The Challenge
As usage increased, some users began experiencing slow application load times and intermittent service failures. ArcGIS Server logs indicated that certain services lacked sufficient resources for consistent performance.
Without monitoring data, the default response would have been to scale up the whole environment or add resources across the board. That can mask the symptoms, but it also raises costs without addressing the root cause. Langan used ArcGIS Monitor to better understand how the city’s ArcGIS Enterprise environment was being used, where pressure was building, and which parts of the architecture were underutilized.
How ArcGIS Monitor Helped Identify the Real Usage Pattern
ArcGIS Monitor provided visibility into historical usage trends across the city’s ArcGIS Enterprise deployment. Langan reviewed key indicators, including CPU and memory utilization, service request volume, response times, and ArcGIS Server instance utilization. These metrics narrowed the investigation and revealed a clear pattern: hosted services were driving nearly all the request volume.
ArcGIS Monitor also showed that memory use on the hosting server typically ran high, while the federated SQL server was getting very little traffic. The original architecture had been built around an expected pattern that no longer matched how the city used the system, and the architecture needed to adapt.
Supporting the Entire Facility Lifecycle
With a clearer picture of the workload, Langan recommended targeted changes to align the city’s AWS infrastructure with how staff used ArcGIS Enterprise. The revised architecture pulled resources back from the underutilized SQL-published services tier and put them behind the hosted services that staff relied on most. Langan scaled down the Amazon RDS PostgreSQL database, deprecated the underutilized federated ArcGIS Server site, and adjusted the hosting architecture to provide more dedicated resources for hosted services.
The biggest change was moving ArcGIS Data Store from the original single-machine base deployment to its own Amazon EC2 instance. The work was scheduled during regular maintenance windows and performed at night to minimize user disruption. The ArcGIS Data Store migration was completed as a hot migration using shared storage, avoiding downtime for that portion of the transition.
Better Performance with Lower Infrastructure Costs
The result was an ArcGIS Enterprise environment that matched the city’s day-to-day workflows. Using ArcGIS Monitor to see how the system was behaving, Langan simultaneously improved performance and reduced infrastructure costs. Rather than scaling everything, the team rightsized the architecture by adding resources where they were needed and cutting them elsewhere.
Chris Dougherty
Sr Solutions Architect Lead
