Yes, the other resources get created fine. As far as I can find, no. I didn't want to open an issue until I got some feedback from the community on whether or not anyone else has seen this behavior (no need to open an issue for something that someone knows was fixed).
If I don't get any more responses in the next few days, I'll open a bug report. I've seen this reproduced on all of my hosts (to the point of having to disable puppetd on some of them, since it was alternately applying correct and incorrect configs), all running 0.24.8. -Jason Andrew Shafer wrote: > > That's a weird one. > > I'm assuming the other resources that are created for the fooClass get > created? > > Possibly an issue with the scope lookup. > > Is there an open issue in redmine? > > > > > On Wed, Oct 14, 2009 at 12:43 PM, Jason Antman <[email protected] > <mailto:[email protected]>> wrote: > > > Good Afternoon, all, > > I believe I posted something on this in the past, but didn't get many > replies. > > I'm trying to manage some slightly different configs for different > classes of hosts within templates using > scope.compiler.classlist.include. Specifically, for example, all > of our > hosts have the same httpd.conf file, with the exception of two hosts > (each of which have a unique class defined in them). > > Something like: > <% if scope.compiler.classlist.include?("fooClass") then %> > foo > bar > baz > <% end %> > > I originally tried this with both the template for httpd.conf and the > template for /etc/sysconfig/iptables. > > On all of the hosts that use these methods, I'm seeing some strange > behavior where intermittently (sometimes as much as 50/50, exactly > 1 on > 1 off) the if statement does not evaluate, and the segment of code > within it doesn't get applied. I've looked through the logs to no > end - > all I can see is that these files flap back and forth endlessly, > removing the generated file, replacing it with a new one, the > replacing > it with the original (the MD5 sums also reveal the A-B-A-B pattern). > > I'm using an external node classifier script, but have been > dumping its' > output to a timestamped log file, and the output is always correct > (the > classes in question always appear in the YAML). > > Any ideas? > > I'm aware of the pervasive theory of using fragment-based file > creation, > but have a few issues: > 1) I very much want to keep all code related to a certain file in one > place (i.e. the iptables module should test for presence of each > service > and then create the iptables file the way it wants to, not including > iptables information in other modules). I'd prefer not to scatter code > related to one concept (iptables, httpd) in different modules. > 2) For some of these files, order is *very* important (we run a ... > complex... iptables ruleset) and, if this code were split across > multiple modules, it would require reading through endless files to > figure out where lines X, Y and Z should go. > > Has anyone else experienced any issues like this? > > Thanks, > Jason Antman > Rutgers University > > PS - All boxes are CentOS 5.3 x86_64 running Puppet 0.24.8 > (0.24.8-4.el5). > > -- > > Linux: The smack in the face that Windows gripers have been > begging for these past 10 years... > > -- > > Jason Antman > www.jasonantman.com <http://www.jasonantman.com> > [email protected] <mailto:[email protected]> > Cell: (201)906-7347 > > Systems Programmer > Rutgers University > OIT Central Systems & Services / NetOps > [email protected] <mailto:[email protected]> > > > > > > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Puppet Users" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/puppet-users?hl=en -~----------~----~----~----~------~----~------~--~---
