Application performance monitoring with Hyperic HQ

System of Logical Defaults

Hyperic makes application metrics available with little to no configuration through the use of a system of logical defaults. Most application monitoring requires a special additional suite of software, which can be tedious to implement. On the other hand, Hyperic HQ out of the box can collect the most relevant performance and service metrics for the most commonly used software and do so with little to no need for manual configuration.

Monitoring Best Practices

A best practice is to monitor only the services and applications that are of importance to the administrator; otherwise, the data becomes less effective under the weight of irrelevant information.

Also, make sure you establish a strategic alerting and escalation policy by configuring alerts that meet multiple conditions. This approach will prevent unnecessary 3am wake-up calls alerting you to a single server that is out of disk space on /tmp. Keep your alerts from "crying wolf."

It takes a significant amount of time and tuning to get a proper monitoring policy that works. Often you'll need a period of at least one or two weeks to narrow down the proper alerting thresholds and identify what you need to monitor. My recommendation is to identify the critical business or organizational processes in order of priority and then map which IT infrastructure services support those processes. From there, you can draft and test a monitoring and alerting policy. In the first phase, alerting thresholds and monitoring metrics should be loose. Over the course of a week or so, identify any unnecessary items, and remove them one at a time.

If the monitoring policy is targeted and well designed, the one or two week grace period provides a fairly simple and smooth transition. Starting with an open, noisy monitoring and alerting policy ensures that monitoring metrics and alerts that would normally be of interest are not accidentally filtered out.

Conclusions

The primary difficulty in performance and enterprise monitoring is extending the solution into all of the particulars of a custom application, the database, and web servers that typically require a suite of expensive, proprietary monitoring tools. Hyperic brings a new era to enterprise monitoring by focusing on deep, application-specific performance metrics and making them virtually free of the need for configuration.

Hyperic HQ is supported by a community. In addition to the built-in support for more than 70 common products, other plugins for additional products and technologies are available from the HQ community site. The HQ server itself is open source, so anyone can modify and contribute to the code; however, most developers contribute to the project by creating plugins for new applications and new server types.

Hyperic HQ does an impressive job of application monitoring and removes the need to spend long hours scripting custom application-specific monitoring tools or to purchase expensive, third-party alternatives.

The Author

Matthew Sacks is a system administrator and technical writer from Los Angeles, CA. He runs a technology blog, The Bitsource, at http://www.thebitsource.com. You may contact him at mailto:matthew@matthewsacks.com.

Buy this article as PDF

Express-Checkout as PDF
Price $2.95
(incl. VAT)

Buy Linux Magazine

SINGLE ISSUES
 
SUBSCRIPTIONS
 
TABLET & SMARTPHONE APPS
Get it on Google Play

US / Canada

Get it on Google Play

UK / Australia

Related content

comments powered by Disqus

Direct Download

Read full article as PDF:

036-040_java_performance.pdf  (527.06 kB)

News