ArcGIS Server is like a GIS office in a box
ArcGIS Server is like a GIS office in a box
In a previous blog we took an alternative perspective on ArcGIS Enterprise and related it to a car to help understand the connections and parts of a base ArcGIS Enterprise deployment. In that article we described ArcGIS Server like the engine of the car, and that description works great in the larger context of ArcGIS Enterprise. But it doesn’t quite explain how ArcGIS Server works internally. So let shift our thinking and re-evaluate ArcGIS Server.
I find it helpful to think of ArcGIS Server like a GIS office in a box. This perspective will allow us to consider different categories of GIS services and explain how those web services in ArcGIS Server work. In this GIS office there are three major roles to examine: the GIS Manager, the office receptionist, and the GIS analysts.
In ArcGIS Server, there are coordinating processes that handle the distribution of requests to web services and the general overall innerworkings of ArcGIS Server. These processes are defined in the ArcGIS software and are not configurable nor managed by users. There are also two web applications used by ArcGIS Server publishers and administrators to configure the GIS web services and the ArcGIS Server site. These applications are the ArcGIS Server Administrator Directory and the ArcGIS Server Manager. In essence they both do the same things, but provide different experiences. The Administrator directory app provides all the tools available to administer and configure ArcGIS Server and its services through a text interface against the REST Admin API. while Server Manager app has about 75% of the functionality and is has GUI. Collectively we can consider these apps and processes as the GIS manager of our office.
In ArcGIS server there is also a specific location that clients can go to get ready made map, vector tile and image service cached tiles. You can think of this as the receptionist’s desk. Most times there’s a person at the desk that can help clients get the tiles needed, this is a process in ArcGIS Server called the tile handler.
Lastly, the ArcGIS server has specific processes that support service requests directly. In our office the GIS analysts are focused on specific resources and can respond to live requests for the resource they are responsible for supplying.
Our GIS office also needs some way to advertise what resources they are making available to the rest of the organization. I like to think of this as a catalog of resources. I can remember the days before ArcGIS Server and portals, where we needed to maintain a basic web page or pdf doc that listed these GIS products. In ArcGIS server there is a web application that comes with the software that provides both a general listing and then much greater in-depth insights to each web service running on the site. This web application is called the ArcGIS Server Services directory.
When our office… or should I say ArcGIS Server... is working, a client can either get an advertised resource from an analyst or from the front desk. The place they go depends on the need of the client and the role the resource plays in the current application talking to the office. They can either get cached or dynamic results from ArcGIS Server. If the application needs a basemap to help provide context in a map it will get cached tiles, but if we need to see the most current state of the data or make some changes to update information managed by the office the client will need to communicate with a dynamic service.
Great! We now understand the structure and hierarchy of our GIS office in a box. In our following articles we discuss what each person does and look at how, as ArcGIS administrators and content creators, we can manage and control the performance of these parts of ArcGIS Server.