Aaron: The application and network service monitor.
News
- 2001-07-05: Looks like Source Forge fixed the CVS and email problems... now if they could get the ssh keys working again....
- 2001-07-04: Made a second pass at the aaron architecture. Much more complete and more well thought out.
- 2001-07-04: Looks like a passwd file is out of sync on the CVS server because it can't look up my UID right now. A correlated problem is CVS commit emails aren't working. I'll post an update when they're working again. I think a security conscious admin was overly zealous with chroot(), but I could be wrong.
- 2001-06-28: First cut at the aaron architecture has been defined.
- 2001-06-27: CVS commit emails should be working
- 2001-06-26: ROADMAP posted, has basic plan of attack
Links
Project Goals
- Aaron will use an Apache style license, not GPL. Being corporation friendly is very important (project sponsored by IT Net App).
- Aaron does:
- check the local host for running programs (httpd)
- check network services (smtp, postgres, http, etc.)
- check network routes (tests for ICMP responses)
- operate in parallel and operates within configuration limits (load, memory usage, etc.). Being light weight when necessary is important. Being able to consume an entire machine's resources is also important.
- have a reaction API (restart service, send email, send SMS message, etc) that is highly configurable through XML.
- support both basic level 4 checks and level 7 service checks (highly customized testing of availability).
- Log via an API making it possible to log to XML, log files, or a database
- collect information through an API making socket, UDP, multicast, and SNMP data collection operations possible
- Aaron will also have an analytical API making it possible for aaron to be aware of "service level agreements." (Ex: if a service is unavailable 0.002% of the time, for example, then alerts will be sent to product manager.)
- Web interfaces will be made available (mod_perl, mod_ruby, PHP, JSP, and servlet) for viewing the host or availability information.
- The configuration and data availability will be setup to allow for data hiding and a permissions model.
- Aaron will be network aware and will be capable of alerting the network administrator in the event of a route failure, programmers in the event of a bad response, system administrators if there isn't a response, and product manager if an SLA has been violated or is near being violated.
- Eventually an AaronD will be created to allow for fail-over and parallel checking of services across multiple services from possibly multiple geographic locations.
Help Needed
Obviously help is needed. I am only one person and can only do so much. Here's my things that I know need to happen:
- A graphics/UI/web designer is needed badly. Basic HTML is about all I like to produce if I'm in a hurry. Can someone please make this beautiful?
- More programmers! The more the merrier!
Feedback
$Id: index.html,v 1.7 2001/07/05 21:37:30 thetitan Exp $