Heroix Monitor: SLA Monitoring Everyone Can Use
In this issue: SLA Monitoring Everyone Can Use
 


SLA Monitoring Everyone Can Use

Service Level Agreements (SLAs) allow you to monitor your critical IT services to ensure that they are in compliance with departmental contractual obligations. Even if you are not subject to service contracts, however, Longitude®’s real time SLA dashboard and reports can help you demonstrate the performance of applications and systems in terms that less technical users and managers will understand.

Longitude includes out-of-the-box SLAs, and you can also define your own. Service Level Agreements are defined in terms of service availability and performance, answering questions such as:

  • "What percentage of the time was the service available?"
  • "How was it performing?"

SLA Dashboards and Reports

Longitude’s SLA Dashboard provides up to the minute displays of SLA compliance using graphs suitable for management reporting, with pinpoint drilldown that supports detailed troubleshooting. Historical reports are also available - with drilldown, of course! 

View SLA Dashboard Samples View SLA Monitoring Dashboard Samples

Defining Meaningful SLAs

Under the covers, SLAs are based on one or more individual metrics (also known as statistics). In addition to monitoring performance counters such as CPU time, SLAs can keep tabs on transactions that you may have configured Longitude to monitor, such as HTTP (web page) availability.

For example, a service condition named “Server Performance” may consist of the metrics CPU Busy, Memory Utilization, and IO Traffic. An “Application Availability” SLA might check whether a certain web page is available and returning expected content, as well as whether a back-end database is functioning properly. Alternatively, service conditions may consist of only one metric; for example, “Network Response” might measure the response time of a ping request to another server on the network. .

Making SLAs Emulate Business Conditions

Longitude provides a realistic picture of SLA performance by taking into account the fact that meeting an SLA normally depends on disparate factors (such as web and database availability in addition to CPU and memory loading), and may operate across a set of computers or disks.

For example, the “Server Performance” condition above could be measured over 3 computers, with a result of “good” if all 3 computers are satisfactory, “degraded” if 2 out of 3 computers are good or degraded (but one may be down completely), etc. This allows SLAs to better emulate business service conditions by taking redundancy into account.


Learn more ...

Next month: Tips for Monitoring Microsoft Exchange