How We Monitor
SiteGuardian has several geographically-dispersed monitoring stations. Once you have configured a monitor, take an HTTP monitor for example, the configuration for that monitor is pushed out to the monitoring stations. Each monitoring station will send out a request to your server and receive a response back. The response received on each monitoring station is sent to a data warehouse where the response is check for alerts that you've previously configured.
Let suppose there are 3 monitoring stations. If the response from each monitoring station has the same alert triggered at the same time period, then a notification(via email or SMS) is sent out to any contacts linked to that monitor. All 3 monitoring stations responses must have the same alert found before a notice to sent out. If only two of the monitoring stations response had the alert, then no notification would be sent out since at least one of the responses was successful.
Informational - The alert is worth capturing and recording, but does not cause the service to slow down or disrupt. An example could be a broken image on your site. It's something that you should be notified about, but it may not be critical that it gets fixed immediately.
Warning - This type of alert usually means that the service may be beginning to have problems. An example would be that your services normally returns a response within 4 seconds, but now it's returning 7 seconds. This may be something that you would want to check into, but has not caused any customer impacts.
Fatal - This alert means that your service is having a problem that may potentially be causing customer impact. This type of alert needs to be looked into immediently. An example would be that that your sites response time is normally 4 seconds, but all of a sudden it's now responsing with 35 second response times. The potential for customers to leave your site is high since they don't want to wait. Another example would be that an application is returning an error instead of the users information. In either case, you would want someone notified about the problem and have it looked into ASAP.
Whenever all the monitoring stations trigger the same alert at the same time period for the same monitor, then it is considered a customer impact and a notification is generated and sent to the contacts. All of the online and email reports will list the number of information, warning, and fatal customer impacts. Alerts that are triggered by one monitoring station and not the others is also logged, but not reported on.
The nice thing about these categories is that any alerts can be assigned to any category. For some users, a broken image may be a fatal error and for others it may not. SiteGuardian allows you to configure alerts and notifications to your business needs.
Alerts and Notifications
Alerts can be assigned to one of three types of categories - Information, Warning, and Fatal. These categories allow you to determine what kind of alerts determine certain actions. Informational - The alert is worth capturing and recording, but does not cause the service to slow down or disrupt. An example could be a broken image on your site. It's something that you should be notified about, but it may not be critical that it gets fixed immediately.
Warning - This type of alert usually means that the service may be beginning to have problems. An example would be that your services normally returns a response within 4 seconds, but now it's returning 7 seconds. This may be something that you would want to check into, but has not caused any customer impacts.
Fatal - This alert means that your service is having a problem that may potentially be causing customer impact. This type of alert needs to be looked into immediently. An example would be that that your sites response time is normally 4 seconds, but all of a sudden it's now responsing with 35 second response times. The potential for customers to leave your site is high since they don't want to wait. Another example would be that an application is returning an error instead of the users information. In either case, you would want someone notified about the problem and have it looked into ASAP.
Whenever all the monitoring stations trigger the same alert at the same time period for the same monitor, then it is considered a customer impact and a notification is generated and sent to the contacts. All of the online and email reports will list the number of information, warning, and fatal customer impacts. Alerts that are triggered by one monitoring station and not the others is also logged, but not reported on.
The nice thing about these categories is that any alerts can be assigned to any category. For some users, a broken image may be a fatal error and for others it may not. SiteGuardian allows you to configure alerts and notifications to your business needs.