What is considered an application outage?
When an application is under duress or is unavailable due to network, application, or other service failures, application servers may not be able to successfully respond to requests from users. When multiple global vantage points fail to receive a response from an application server after 5 seconds or receive an error message, the application is considered to be unavailable and experiencing an “outage” event from the standpoint of users in impacted locations.
Application outages depicted on the map may be caused by an issue with the application provider (either software or infrastructure related), or may be due to a third-party outside of the provider preventing reachability of the application. The outage map does not report on performance-related issues within applications. For example, an application may be running very slowly or some or all of its features may be unusable, but its servers are reachable and responsive. In that case, it would not be categorized as an “outage” incident and would not appear on the map.
Details on the underlying issue for each detected application outage are available via ThousandEyes Internet Insights subscription service.
What thresholds are in place for reporting an application outage on the Outage Map?
Outages that last less than five minutes and involve fewer than 5 servers are not reported on the Outage Map. They are, however, reported in the paid ThousandEyes Internet Insights subscription service. Learn more about Internet Insights here.
Why don’t I see an application outage that has been reported by users and/or the application provider?
There are a few reasons this might happen. First, it may be that the application is not one of the approximately 75 popular SaaS providers currently being tracked. Second, it may not have met the thresholds defined above in terms of duration or number of servers affected. Third, the outage could be impacting the useability of an application or some application service. For example, an authentication service issue may prevent users from logging in or prevent some action within the application due to a disruption in the application’s process flow. These issues would not be reported on the outage map. Finally, it may be that users are mis-reporting an application outage when that is a symptom, rather than the cause. For example, an ISP in the service delivery path may be experiencing an issue that is preventing users from accessing an application. A user may report the application to be impacted, when in reality, the application is working fine, it’s the service provider delivering the application that is experiencing an issue.
How can I tell on the map if a network is what is impacting the application experience?
This capability is only available for ThousandEyes customers. Learn more about Internet Insights here.
How do I understand where an outage is taking place?
Application outages are highlighted on the world map using purple dots. Hover over each dot to determine which applications are experiencing outages in those regions.
How do I understand if the outage is significant or not?
Significance is usually determined by length of time and number of servers / locations impacted. The higher those numbers are, the more users that are potentially affected by the disruption.
Can I submit a request for an application I care about to be tracked?
ThousandEyes customers can submit a request via support, using the ‘I wish’ feature in the product, and directly to your account manager. We are always interested in hearing about additional applications that our customers and visitors would benefit from having available in the Internet Insights catalog.
What can I do with this data?
Share the outage event with others who might be impacted by the outage using the embedded “Share” button on each outage card. You can alert the provider to the outage if you’re unsure they’re already aware by sharing a direct link to the map. Customers of ThousandEyes Internet Insights can be automatically alerted of outage events, enabling them to proactively alert their workforce of any business-disrupting outages to avoid an unnecessary flood of helpdesk tickets.