January 2008: Hyperic HQ Enterprise Monitoring
Project name: Hyperic HQ Enterprise Monitoring
Date founded/started: 2006-07-17 (registered on SourceForge.net)
Project page: http://sourceforge.net/projects/hyperic-hq
Description of project:
Hyperic HQ – open source systems management, Apache management, Tomcat management, LAMP management, WebLogic management, virtualization management, VMware management, WebSphere management, Xen management, JBoss management, and WAMP management.
- Database Environment : JDBC, MySQL, Oracle, PostgreSQL (pgsql)
- Development Status : 5 – Production/Stable
- Intended Audience : Developers, System Administrators
- License : GNU General Public License (GPL)
- Operating System : OS Independent (Written in an interpreted language)
- Topic : Enterprise, Monitoring, Systems Administration
- User Interface : Plugins, Command-line, Web-based
How would you describe Hyperic HQ?
Hyperic HQ is open source systems management software designed for complex Web infrastructures of all sizes. Its ability to automatically discover, integrate, and manage new technologies enables companies to easily and effectively maintain business-critical services. Hyperic HQ’s enterprise-ready multiplatform approach to monitoring and management helps pinpoint, correct, and prevent problems at every major layer, including hardware, networks, virtualization, and applications.
Why and how did you get started?
HQ had several starts. First it started as a project within Covalent designed to manage and monitor the Web infrastructure of Apache, Tomcat, and databases like PostgreSQL and MySQL. Its first name was actually Covalent Application Manager (CAM). Eventually, Covalent decided to focus on other areas, and we decided to take CAM and form Hyperic. Hyperic continued investment and working with about a dozen customers to build out what HQ would eventually become. Part of that work included making HQ open source. On July 17, 2006, we finally achieved that milestone and launched the Hyperic HQ project on SourceForge.net.
What is the software’s intended audience?
We built HQ to help operations personnel better monitor and manage the general complexity and noise associated with Web deployments. It’s designed to help organizations keep up with change, provide broad product support, and set alert conditions and control actions that help operations staff focus on real problems and weed out false alerts fast.
How many people do you believe are using your software?
We know of somewhere around 1,500 real deployments that have made their project success known to us, but since it is open source, not everyone has to interact directly with Hyperic, so the real number is somewhere north of that.
What are a couple of notable examples of how people are using your software?
Customers include CNET Networks, hi5 Networks, and eHarmony.com. HQ helps many of our users run a complete 24×7 operations with no night staff — just Hyperic. Most of our users provide e-commerce and other Web-based businesses. Frequently these folks have a relatively small yet senior team of operations engineers who support a business that can’t experience downtime or any reduced availability. For the best examples, check out our case studies, where our users explain their implementation strategy and their ROI from using Hyperic.
What gave you an indication that your project was becoming successful?
One was the first day we saw Hyperic HQ as a skill required in a job description. Mostly, however, we are reminded of its success every day simply by participating in our forums and seeing community members help each other.
What has been your biggest surprise?
We first expected to rely mostly on search and our own site to encourage downloads. As our project downloads increased, we found that a tremendous amount of our user base finds us directly on SourceForge.net.
What has been your biggest challenge?
Managing the wide array of technologies Hyperic supports. We autodiscover and manage more than 70 technologies across nine different operating systems. Making sure we stay current and provide the deepest level of monitoring and control capabilities for all of those is a constant challenge. Thankfully, our community has begun to step up and help us here — developing and sometimes testing new versions with us.
Why do you think your project has been so well received?
We solve a hard problem for people who usually just build their own solutions or suffer through a lot of pain. Autodiscovery gives instant gratification to new users, and ensures they are aware of all changes in their environment from then on. This is huge, because change usually accounts for about 70% of what went wrong when a problem occurs, so this really keeps them ahead of issues before they occur.
Where do you see your project going?
Currently Hyperic HQ does a good job of discovering and managing Web infrastructure as soon as it’s installed. Where we would like to take it is to more custom-fit that experience to the types of Web infrastructure you are installing — providing what we call the zero-config experience, where you download the software and it is smart enough to set itself up, establish best practices for monitoring and alert schemes, and continue to keep up with changes in your environment.
What’s on your project wish list?
Some current priorities we are working on now are:
- Scalability — 10,000 hosts on a single HQ installation.
- Enhancing the UI to provide a more tailored, flexible, and customizable user experience.
- Better resource model to allow better dependency checking for alert definitions.
- Provide a roll-up capability in HQ that aggregates performance data from several HQ installations in a regional deployment scenario.
- Provide better support for scheduled downtime.
What are you most proud of?
The whole community that is Hyperic. The team employed by Hyperic came from some really impressive places — Netscape, Oracle, Sun, Wily, Apple, and more. Our users are equally impressive — including a dozen or so of the top Web sites in the world, huge government installations, top-tier financial institutions, airlines, breweries, healthcare companies, and universities.
If you could change something about the project, what would it be?
Shorter release cycles and more input from the community on betas. These are things we are actively working on.
How do you coordinate the project?
We use agile development and come together once a day to discuss tasks, dependencies, and how to break down any roadblocks to completing our development tasks. Bugs are verified through QA and assigned by the project lead to the various developers. QA has a formal set of requirements for testing functionality and performance at every stage. Recently, we also built a performance engineering team that designed “super-tests” to simulate extreme data throughput with the application. This really helped the 3.2 dev process enhance scalability.
Do you work on the project full-time, or do you have another job?
All except one committer is employed by Hyperic full-time.
What is your development environment like?
We are a Mac shop. All development happens using Intelli-J and Eclipse, and we use a lot of VMware and Parallels for alternate OS and browser development and testing. There are some other server types around for common use for developers, but all developers have a Mac.
January 2008 (planned): Hyperic HQ 3.2
- New diagnostic capabilities
- Ability to run Nagios natively, with new UI
- MySQL database support
- Improved scalability
August 2007: Hyperic HQ 3.1
- Alert center
- Localizeable UI
- Contextual help
- 18 new plugins
January 2007: Hyperic HQ 3.0
- Operations dashboard
- RSS feeds for alerts
- New UI
- Control actions
July 2006: Hyperic HQ 2.7
- First open source release
How can others contribute?
Check out our previous projects of the month.