| NOTE: This site is no longer maintained. Upgrade to ipMonitor v9 here. |
| Every organization relies on certain key systems and applications that are vital to its day-to-day business and long-term prosperity. |  |
 |
Infrastructure equipment and applications that serve employees, partners and customers -- such as SQL servers, commerce solutions and dynamic web applications -- are required to be available twenty-four hours a day, seven days a week. Outages could have disastrous effects on a corporation's ability to conduct business, its reputation and its profit margins. But ensuring that these mission-critical resources are available around the clock is only part of the solution.
Synthetic transactions used to test the performance of your systems and applications should be part of your end-to-end network monitoring strategy. Monitoring solutions that support transaction-level testing are able to do more than simply test whether your key resources are available. They are also able to ensure that these mission-critical systems and applications are always operating at peak performance. Synthetic transactions accurately simulate the user experience, making it possible for the monitoring software to perform multiple transaction tests, analyze test results and measure response time.
Different network monitoring solutions handle transaction-level monitoring in various ways. Some require you to install agents on each machine in order to capture the necessary performance data. Others are able to collect performance metrics across your network from one central location, without requiring an agent on every system. Ultimately, what all monitoring packages that handle synthetic transactions have in common is the ability to monitor the Quality of Service experienced by end users.
Targeted threshold values can be set for each type of resource being monitored. When these predefined values are exceeded, the appropriate system administrator will be alerted. For example, Microsoft® Outlook® depends on multiple Microsoft Exchange Services to categorize, queue, route and deliver email, as well as to authenticate a user. It requires all of these components to function correctly in order to accept and deliver a single email. If any one of these were to become degraded or fail altogether, mail service for an entire organization could be disrupted. If the organization had a monitoring system in place but had only set it up to monitor availability, finding the root cause of the performance problem could take hours. However, with transaction-level monitoring properly configured, an administrator would be notified of the exact mail subsystem causing the issue.
Once your monitoring system has been in place for a period of time, you can use the metrics collected to identify trends and build a performance profile by taking advantage of the monitoring software's reporting capabilities. This allows you to notice, for example, if your mail server responds slowly during a specific time of day due to heavy user load. Using the performance data gathered, you can also establish whether a particular mail server subsystem is often the source of the problem.
In addition, many network monitoring products support the ability to display test results in real time, which makes it possible to determine the overall health of your mission-critical systems and applications at a glance. Performance problems can then be dealt with as they happen, before they can escalate to critical levels.
Use a network monitoring software package capable of performing transaction-level tests to monitor the following resources in your organization:
- DNS Servers - Test the ability of a Primary and Secondary DNS server to respond to a record query, as well as its level of responsiveness.
- Mail Servers (IMAP4, POP3, MAPI) - Test the ability of your SMTP server to receive and distribute email, as well as the ability of your end-users to log in from an IMAP4, POP3 or MAPI-enabled email client and manage their email.
- SNMP-enabled devices, services and applications - Retrieve a numeric or textual value from an SNMP agent and then test it against rules you define. This process can be used to identify ICMP flood attacks, trap spikes in network traffic, identify packet failures, ensure a minimum level of network availability, monitor environmental conditions in server rooms and cabinets and identify hardware problems in advance of costly failures.
- FTP servers - Test a FTP server's ability to accept incoming sessions, process user logons and then transmit the requested file.
- Web servers - Test a web server's ability to accept incoming sessions and transmit a requested resource, such as a web page, or the results of a CGI script.
- SQL servers - Monitor a SQL Server from the end-user perspective by using synthetic transactions to test login ability, perform a query, retrieve data and analyze results for correctness.
Transaction-level monitoring helps ensure that your mission-critical systems and applications are always available, responding as required and achieving the maximum performance and reliability necessary to deliver optimal Quality of Service to your end-users.
For information on other features and concepts related to those discussed in this article, refer to the following ipMonitor resources:
< Back
Last Updated: December 13, 2006 | What did you think of this topic?