Hi all,
For those of you following along at home, you may have noticed that
Heka's `pipeline` package was growing pretty unwieldy. Not only did it
contain the code for Heka's core processing engine, it also contained
the full implementation for every plugin that ships as part of the Heka
core. This wasn't a problem when we were starting out, but as the corpus
of plugins has grown the package was growing increasingly unwieldy,
making it harder to get oriented, find code, and avoid intra-package
namespace collisions.
I'm about to land a change on the dev branch that cleans this up,
creating a `plugins` package containing a large number of sub-packages,
organizing the plugin implementations into bite sized logical groups.
This should have no impact on Heka's behavior, but it does mean that the
scope of a few functions and attributes have changed (from package-only
to public), and lots of code that used to be in the pipeline package is
now somewhere else entirely.
If you're tracking dev but you're just using Heka, this should have no
impact, other than that you'll want to delete the entire `build`
directory and re-run `source build.sh` (or `build.bat` on Windows). If
you've developed your own plugins, you may have some breakage when you
try to use them with the newer code. Luckily, fixing your plugins to
work w/ the new code layout should be pretty easy, speaking as someone
who's just done this a couple of dozen times over. You can peek at
existing plugin code in its new location for tips, and/or feel free to
ask for help either here on the list or in #heka on irc.mozilla.org.
Thanks!
-r
_______________________________________________
Heka mailing list
[email protected]
https://mail.mozilla.org/listinfo/heka
