We have been thinking about how we build and release software, and have decided 
that we will try implementing a more "agile" based approach than we have to 
date.  Specifically, this means is that we will be releasing the Zenoss Core 
software every 30 days.  The term for these iterations is "sprint".  Our goal 
is to release something from these sprints, that is usable and gives us 
(internally) and you (the community) a chance to provide feedback on the new 
features that have been added during the sprint.  Often features will be 
partially completed, but even in that case things should be generally working.

Releases will be versioned using our current alpha numbering scheme.  So sprint 
#1 for this release will be 2.1.70 we will increase the number by one for each 
sprint.  When we are at release candidate level the version will start at 
x.x.90.  This will usually be the last sprint before the final release.

We are releasing our first 30 day sprint today, version 2.1.70.  A detailed 
list of changes with associated ticket numbers is included in the release notes.

At a high level here are the major features:

1. Inclusion of python 2.4, which is required by Zope, within the Zenoss 
release.  This allows us to build Zenoss on platforms that don't have 2.4, such 
as EL4, FC8, SLES 10.3, Ubuntu 7.04 and maybe others.

2. Daemons now have a watchdog process.  The watchdog tracks the health of each 
daemon and if it stops responding the process is restarted by the watchdog.  
This is how Zeo and Zope have been working.  It is also a band-aid that we hope 
will make the problems with the zenwin* processes hanging better.

3. Refactor of how configurations are sent to the collectors from zenhub.  
Please keep an eye out for any issues with collectors that might be caused by 
this change.  Things to look for include collectors failing to get their 
configurations and events not clearing properly.  The change will help us 
maintain the zenhub to collector communications in the long run.

4. Collectors now log statistics on their own health.  Graphs for these are on 
the Performance tab on each performance configuration object.  We are tracking 
things like monitor cycle time and number of data points collected.  This 
should help when tracking down performance problems with the collectors.

-EAD

Erik A. Dahl
Co-Founder and CTO, Zenoss Inc.
http://www.zenoss.com




-------------------- m2f --------------------

Read this topic online here:
http://community.zenoss.com/forums/viewtopic.php?p=15069#15069

-------------------- m2f --------------------



_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users

Reply via email to