Some linting fixes and resolves joining wan not actually joining the wan #66
Conversation
|
Whoa! |
|
Can you paste the actual output of It's certainly possible we've been wrong this whole time, but I would like to see your output to confirm. |
|
Here are two entries from two different clusters: web1east.east 10.30.0.13:8302 alive role=consul,dc=east,vsn=2,vsn_min=1,vsn_max=2,build=0.4.1:,port=8300,expect=3 Essentially if the data-center that the server is in isn't attached to the remote data-center there will be no entries returned by the -vP grep which will then generate an exit 1 code which runs the exec command, thus is must be an 'unless' statement. Trust me, I tested the exit results thoroughly. |
|
Ok. Can you rebase without the init style changes and only have the wan-join thing? |
|
I can remove those but I suggest leaving them. They are lint fixes as you are not supposed to use selectors inside of resource definitions. I can link to the puppet documentation on this point if you'd like. |
|
Ah nevermind. Ship! |
|
Will you be merging this or is there changes still outstanding you'd like me to make? |
|
Yes, please rebase on master so I can click the merge button. |
|
@EvanKrall do you have a second opinion? How was wan joining working for us this time? |
|
I haven't bumped our puppet-consul module in a long time (since before wan joining was added), and our profile_consul handles wan joining. This lg2m. |
* https://github.com/solarkennedy/puppet-consul: Add in a couple of missing test assertions Add support for providing watches/checks/services via hiera into the consul class Update install.pp Adding ability to disable managing of the service removed redundant facts in spec tests Fixed tests Better version checks for OS Conflicts: manifests/run_service.pp
|
Okay, should be good to go now. |
Some linting fixes and resolves joining wan not actually joining the wan
|
Whew! Thanks. |
No description provided.