Design choices and considerations

The design of this reference architecture assumes that the Network Management System is a business and/or mission-critical enterprise system, where requirements include:

  • Minimal system downtime due to expected or unexpected events
  • Excellent service performance, without substantial lag that would hinder end-user productivity.
  • An efficient, effective, and overall positive end-user experience.

To achieve these goals, the design needs to meet certain criteria within each architecture pillar. Some of the recommended practices associated with each pillar are outlined below, but they do not represent a complete set of architectural considerations. Refer to the data editing and management system pattern’s considerations for more information.

Performance and scalability

In terms of performance and scalability, this architecture aims to optimize the overall experience users have with the system, while responding to evolving workload demands. A Network Management System should deliver editing experiences with consistent performance metrics, to create a positive end-user experience that increases end-user efficiency. In addition to the performance improvement practices outlined below, relational database performance management is also a major factor in the overall performance of your Network Management System.

Workload separation

Workload separation is a design approach focused on optimal distribution of compute resources. For example, some editing requests in a Network Management System may take longer to process than standard map requests, so editing workloads may benefit from separate, dedicated compute resources (such as an ArcGIS GIS Server site). This approach of workload separation helps separate long-running requests from shorter requests so that editors have dedicated resources and viewers are not impacted by long transactions. If this separation is implemented, the system performance for both groups is likely to improve as resource contention is reduced and the system is more easily scaled - resources can be added to either server site to scale horizontally or vertically. Workload separation can take several forms:

  • By component Separating components onto different virtual machines or compute infrastructure ensures that individual components are not contending for system resources. While ArcGIS Enterprise supports installing and configuring multiple components on a single system, this is generally not recommended in well-architected production systems.

  • By service type Another workload separation approach is applied to this architecture within the ArcGIS Server components – workload separation by service type. Separate GIS Server sites support Utility Network workloads and hosted service or mapping workloads.

Co-location

Co-location is a design approach where system components are deployed to the same data center, in the same sub-network, which helps reduce network latency by reducing the communication distance across the network. In general, network latency is more impactful to end-user experience than network bandwidth for common GIS operations. Another consideration in this area is the location of user and client machines – if a user has a high latency connection, the co-location of system components is unlikely to improve their experience of working with the system. In some cases, use of thin clients or remote access may be preferable to physical hardware that connects over a slow or overloaded network.

Reliability

Reliability ensures your system provides the level of service required by the business, as well as your customers and stakeholders. As a business or mission-critical enterprise system, Network Management Systems always require backups of the data, and often require backups of system components. They may also require a high availability configuration to achieve higher levels of uptime.

Backups

For enterprise systems with availability expectations, requirements, or commitments, a clearly defined, actionable, and well-tested backup approach is critical. With Network Management Systems, data-level backups of the ArcGIS Utility Network are essential (at a minimum). Depending on an organization’s requirements, backups of other system components may also be needed. Refer to backups and disaster recovery for more information on backup strategies and methods.

High availability

High availability is a design approach that aims to build the system to meet a prearranged level of operational performance over a specific period. A highly available system needs redundancy, system monitoring, and automation commensurate with the target service level agreement (SLA). Redundancy might include disparate components such as network connectivity, power reliability, data center cooling, and access to staff with the skills to maintain the system. Automation might be designed to take action based on monitoring to avoid outages. For more information, see configuring highly available ArcGIS Enterprise components.

Note:

Keep in mind that high availability configurations significantly increase infrastructure and operational costs of the system, and requires specialized skills to ensure its success. High availability designs require an operational commitment across people, process, technology, and governance.

Observability

Observability provides visibility into the system, enabling operations staff and other technical roles to keep the system running in a healthy, steady state. Monitoring of system availability, performance, and usage is critical to a Network Management System. In addition to monitoring the ArcGIS Enterprise software, it is important to monitor all supporting components and infrastructure, such as the Windows or Linux operating system, databases and other data stores, compute, network, security perimeter and any other relevant components.

Monitoring

Any organization must have an enterprise IT monitoring and response framework in order to successfully build and operate enterprise systems. Proactive monitoring of systems is as important as reactive problem-solving, and effective capture of telemetry provides awareness of the system at any given time and identifies trending system behaviors.

ArcGIS Enterprise on Windows/Linux can be observed in a variety of ways including server logs and server statistics. In addition to monitoring the ArcGIS Enterprise software, it is important to monitor all supporting components and infrastructure such as the Windows or Linux operating system, databases and other data stores, as well as compute, network, security, and other infrastructure.

In the reference architecture, the telemetry capture mechanism is represented by ArcGIS Monitor. Some important system characteristics captured in this way include:

  • CPU usage
  • RAM consumption
  • Disk activity
  • Network activity
  • ArcSOC usage in GIS Server sites

Capturing Telemetry

A key aspect of observability is the use of telemetry - data or information that represent real user activity on a system. Capturing telemetry across all components of the design, including desktop client machines, is critical to understand the performance and utilization of the system, with the overall intent of identifying bottlenecks and opportunities for optimizing the system.

It’s important to note that telemetry outputs do not necessarily capture the full user experience. Consider engaging with end-users of applications (such as those using desktop, web, or mobile apps) to observe their experience performing their workflows in addition to telemetry capture.

Security

Security protects your systems and information. Security design considerations for a Network Management System are closely aligned to the system pattern security requirements, including important considerations for user authentication, system authorization, data and access control, and auditing of user activity and system configuration changes.

Integration

Integration connects this system with other systems for delivering enterprise services and amplifying organizational productivity. A Network Management System typically needs to accommodate data exchange and alignment with other systems like Enterprise Asset Management (EAM), Customer Relationship Management (CRM), and Advanced Distribution Management (ADMS) systems. Integration requirements for a Network Management System are closely aligned to the data editing and management system pattern.

Automation

Automation aims to reduce effort spent on manual deployment and operational tasks, leading to increased operational efficiency as well as reduction in human-introduced system anomalies. Automation requirements for a Network Management System are closely aligned to the data editing and management system pattern, and include practices like:

  • Extensive workflow automation, like with Tasks in ArcGIS Pro.
  • Use of Python scripting to automate repeatable data management tasks, like QA/QC checks.
  • System administration automation, including software deployment, use of infrastructure as code (IaC), and a DevOps approach to any custom application development.

Physical design considerations

Esri offers system architecture design services should you need help determining all of the different factors relating to your organization’s physical design,such as networking, storage, system environments, and sizing.

Additional resources related to physical design include:

Top