[chef-dev] 3/7 Opscode Ticket Triage

[Bryan McLellan < >]

It is sometimes too easy to forget how important Ohai is. We've been
adding a bunch of additional tests around the Linux network plugin
with iproute and we should see some side-effect features like vlan
attributes. Many thanks to Laurent for having a bizarre desktop system
and being willing to test against it. I'll keep plugging away at it,
keep an eye on master or wait for the next RC email.

I've usually been merging triaged tickets when we make a release, but
we're going to try dedicating some time every Thursday to merging and
seeing if that helps those who work off of master. We will keep
running spec tests before we merge, but master will continue to be
otherwise untested. Stephen Delano wants a synonym that starts with T
for merging, but "Triage Tuesday" is on Wednesday these days, so he is
really just trolling.

To merge:
  CHEF-2956 - Update gemspec to use latest RSpec
  CHEF-2852 - can't modify frozen string
  CHEF-2599 - remote_directory fails trying to purge a non-empty directory
  OHAI-296 - Adjust OpenVZ to support new proc entries
  CHEF-2922 - ability to reload a recipe, for use in shef
  COOK-1084 - Many cookbooks broken on FreeBSD < 8.2
  COOK-1074 - freebsd cookbook that abstracts away platform peculiarities
  COOK-1043 - Bluepill cookbook fails on OS X because it tries to use
'root' group
  COOK-1053 - Add :src_url attribute to nginx cookbook
  COOK-1062 - Databases: Postgres exists should close connection
  COOK-1077 - ssh_known_hosts should allow adding arbitrary host keys
from a data bag
  COOK-1080 - Fixes for centos/rhel boxen and pdns::recursor cookbook

Other:
  CHEF-2964 - FreeBSD service provider is too sensitive to the current state
  Reopened - Provide tests and use shellout
  CHEF-2938 - Add a luarocks provider
  Reopened - Use shellout and rebase instead of merge
  CHEF-2959 - Allow downgrade of packages via the
Chef::Provider::Package::Apt provider.
  Reopened - Resource specific attribute that may not be needed
  CHEF-2958 - gem_package install fails when gem_binary is specified
and source is a local file
  Open - needs a test, I need to follow up on CCLA
  CHEF-2944 - Chef client wedges in sleep state under certain
conditions (likely when chef server barfs on a response).
  Commented - Suggested using iptables to recreate the server wedge.
  COOK-1009 - mysql gem not available within chef run on centos after
running mysql::client recipe on node
  Reopened - What is the least surprise, using system libraries or
rubygems? Does this change by distribution? opinions?
  COOK-970 - database::ebs_volume fails with "Chef::Exceptions::Mount:
Device /dev/sdi does not exist" on Ubuntu 11.04
  Further review: Check the device names for Ubuntu 10.04 to 11.10.
  COOK-1002 - chef-client service is not started for init_style = init
  What is the use case where the service is not started in the first
place? the template resources that get dropped off should take care of
that on the initial run.
  COOK-671 - python_pip resource's call to grep captures modules with
similar names
  Reopen: Used patch from COOK-1036 since this ticket wasn't in
resolve state and was missed in triage. Reopening in case author wants
to refactor grep to the Ruby code
  COOK-974 - chef-server::apache-proxy should make it easy to redirect
http://<host_name>/ to https://<host_name:444 to access the webui
  Reopen: Consistently use the same attributes from the other recipe
as the other ticket's change is in the released cookbook
  COOK-1047 - Option to include system site-packages in python virtualenv
  Reopened - Non-existent attribute used
  COOK-1048 - installation of pip does not honor selected python version
  Reopened - Needs to be rebased against master
  COOK-1066 - chef-client cookbook should start chef after enabling the 
service
  Closed - Duplicates COOK-1002
  COOK-1069 - Typo on chef_handler README
  Commented - Pull request against the wrong repository
  COOK-1081 - Added support for centos/rhel systems which use
/etc/sysconfig/memcached in memcached cookbook
  Reopened - Appears not to write memcached.conf on Redhat

-- 
Bryan McLellan | opscode | technical program manager, open source
(c) 206.607.7108 | (t) @btmspox | (b) http://blog.loftninjas.org