Multiple metrics are used to define the health of your enterprise GIS. ArcGIS Monitor was created for administrators to quickly identify and deal with problems with ArcGIS, whether they are in the software or elsewhere in the stack. To properly diagnose if your GIS systems are healthy, administrators need to be aware of when alerts happen and their criticality as well as understand their root cause.

Alerts

When the system is not healthy, there might several hundreds of alerts in your dashboard. To help better analyze the alerts, ArcGIS Monitor divides them into three categories – critical, warning and info. Addressing critical alerts is top priority for every administrator as they impact availability and the smooth running of your enterprise GIS. Warning alerts indicate resources running low – memory, disk, CPU or network bandwidth. Info alerts are logs that are informational for the administrator.

There are several options to investigate errors – you can view when the error occurred, parse log entries for that time or click on the included log errors links and admin URLs to check site details.

Root Cause Analysis

While administrators need to know when alerts happen, it is also essential that they understand the root cause, the source and the impact, of a problem. For example, an outage of an ArcGIS Data Store impacts all of the tiers above it. The source, in this case, would be the ADS and the impact would be ArcGIS Server and portal sites affected by the outage.

One of the common root causes is system overload. When the system receives loads exceeding its capacity, this results in excessive resource utilization such as 100% CPU, zero free memory, or zero idle disk. This, in turn, lowers performance, causes time outs and impacts overall stability of the enterprise implementation. 

Another common root cause is system bottlenecks, which impact performance and stability while the resource utilization is low. Bottlenecks manifest during increased user load such as the above case.

Lastly, unstable infrastructure is another cause to look for. Restarting services, changing permissions, expired passwords or virtualization overallocation can impact system stability. Examples include unexpected processes consuming memory, CPU usage spikes, stopped ArcGIS Server services, reboot conditions and databases not running.

ArcGIS Monitor provides reports that speak the language of administrators, enabling easy diagnosis of the health of your enterprise GIS, and manage GIS hardware and infrastructure needs. Monitor shows you where the issues are through quantifiable key performance indicators and metrics.

This video demonstrates the above mentioned key features.

Are you attending the Esri Partner Conference and Developer Summit in Palm Springs, CA next week? Stop by the ArcGIS Monitor stations inside ArcGIS Enterprise area at the expo. Our experts are available to demonstrate the product's capabilities and how your organization can benefit from using ArcGIS Monitor. 

Don't forget to attend the ArcGIS Monitor: An Introduction technical sessions to learn more. 

• Thursday, March 4th at 2.30 PM (Catalina/Madera room)
• Friday, March 5th at 1.00 PM (Mesquite C)

ArcGIS Monitor 10.6.1 includes new features and enhancements to make you more productive. Here’s what you can do:

Alerts filtering with links

Quickly filter and sort open or closed alerts with links to charts, statistics, and in some cases log details, or the source url. With one click, you can access severe or warning log entries for ArcGIS Server.

Explore Demo Site > Alerts tab

View current status icons

Use high-level status icons to quickly communicate collection status for managers. A count of healthy and unhealthy collections and count of alerts also help summarize the status more efficiently.

Explore Demo Site > Home tab

View availability status icons

Use high-level icons to quickly communicate the count of collections where availability is less than 99%. This allows administrators to focus on areas where availability was lowest for a time range.

Explore Demo Site > Availability tab

Monitor ArcGIS GeoAnalytics Server

When using GeoAnalytics Server, view new counters including status, cores, and memory for Workers in the GeoAnalytics tab.

View Health Check counters

Quickly check the Health Check counter for a site obtained directly from the ArcGIS REST API. A Health Check counter was added for both ArcGIS and Portal counter types in the Categories > ArcGIS report. A value of zero indicates the site level url is healthy and a value of 1 indicates the site is unhealthy. The Health Check counter is separate from ArcGIS Server logs. A site url may be healthy but have service level errors in the logs.

Use Windows authentication for ArcGIS, Portal, and Http counter types

Conveniently use credentials of the Windows administrative users in the ArcGIS Server built-in store or Active Directory to authenticate for ArcGIS, Portal, and Http in both federated and non-federated environments. Secured ArcGIS Server services requiring a token can also use Windows authentication if the Windows user that is running the ArcGIS Monitor Service has authorization to access the target site or service. Either ArcGIS Server authentication or Web Tier authentication can be configured.

View International dates and number formats

Internationalization support was added for date, time, and number formats to improve the usability for international environments. Unicode support was added for the installation, configuration, and data viewing of non-English usernames, passwords, labels, and input text.

Lastly, 6 new videos are available for ArcGIS Monitor 10.6.1:

1. ArcGIS Monitor Demo for Managers
2. ArcGIS Monitor Demo for Administrators
3. Understanding Availability
4. Add Portal
5. Add ArcGIS Server
6. Add Http

Questions? Comment below and we'll get back to you soon!

A Proactive Approach to GIS Implementations

Williams, an energy infrastructure company, owns and operates midstream gathering and processing assets as well as interstate natural gas pipelines. Founded in 1908, Williams employs more than 5,600 people. With headquarters in Tulsa, Oklahoma, Williams also operates major offices in Houston, Texas; Pittsburgh, Pennsylvania; Oklahoma City; and Salt Lake City, Utah.

The geographic information system (GIS) environment at Williams includes a Citrix environment with an average of 120 ArcGIS Desktop users and supports 63 development, QA, and production servers (mostly virtual). The clustered ArcGIS Server environment hosts services and 164 maps. A federated ArcGIS portal runs 238 services, 199 maps, 30 feature services, and accommodates 1,800 users. With so many workers depending on GIS maps and services, Williams needed to be able to proactively monitor its implementation.

"Anytime a web map or map service wasn't working correctly, we would have to research what's causing the issue," said Steve McCarthy, a systems architect at Williams. "We'd start by looking at the web app, then maybe end up having to dig into the code to figure out what's not working right. This could take several hours or days."

McCarthy and his team would have to take time away from projects to determine the cause of the problem—whether it be a broken link, a bad index, or an underpowered service. He and his team needed to be able to identify problems or potential problems and reliably run multiple applications in their GIS environment. This would allow qualified staff to focus on creating new applications or fine-tuning current applications rather than troubleshooting.

Williams implemented ArcGIS Monitor, a tool uniquely tailored to monitor the health of ArcGIS implementations. ArcGIS Monitor provides insightful information about system usage and performance via noninvasive monitoring of enterprise GIS and IT infrastructure including databases, networks, and GIS software. The solution provides timely detection of potential and existing system infrastructure and operational problems to facilitate a rapid resolution. It also provides actionable reports and quantifiable metrics to improve communications among GIS and IT staff, business owners, and senior management.

Since implementing ArcGIS Monitor, Williams has reduced the amount of time spent on troubleshooting and response. The company has also been able to save the cost of extra resources by making those purchases only when proved to be necessary.

"ArcGIS Monitor provides reports that are very helpful," McCarthy said. "It lets us see where there is heavy traffic. The Excel reports show` me where I might have issues such as broken links or data change by hour or over a period of time."

With a better grasp of available resources, McCarthy and his team can quickly determine how much growing room they have and when they may need to add resources such as extra servers, storage, memory, or CPUs. His team uses ArcGIS Monitor reports to justify the use of resources and the need for more.

McCarthy can fine-tune the ArcGIS Server environment at Williams to cater to whatever services are popular. He uses feedback from ArcGIS Monitor to increase resources based on heavy activity or decrease them based on low activity. While McCarthy elected to receive alerts from ArcGIS Monitor when the server reaches 75 percent capacity, users can set alerts to any desired size or percentage of capacity. ArcGIS Monitor also provides hourly reports on all systems. The team at Williams makes these published reports available to developers and users who need relevant information.

"We've gone from responding to issues to proactive, high-performance tuning," McCarthy said. "Now, we don't go buy extra resources unless we can justify it. In the past, we would only be able to assume we need more memory or more servers. ArcGIS Monitor helps build confidence in our GIS because we have very little downtime."

For more information contact Esri at arcgismonitor@esri.com.