Hi!
as some of you may know Apache Bigtop has developed
quite a bit of Puppet code that aims at deploying all the
projects in Bigtop distribution of Hadoop ecosystem
projects. I am now looking into pairing it up with Whirr's
puppet service and it seems that I'm running into
an issue with not having access to the cluster topology
data
What I am saying is that it seems like there's no way for
me of exposing *all* of the roles/instances data to the puppet
code that runs on each node. What I really would like to see
is a sort of data that ends up in my
~/.whirr/<name>/instances
to be available via extlookup/hiera on every node where puppet
runs.
Thoughts on how this can be accomplished? Btw, I'm fine with
augmenting the current implementation of puppet service in
Whirr, but I thought I'd ask first ;-)
Thanks,
Roman.
P.S. Those who are curious what Bigtop needs this for:
https://git-wip-us.apache.org/repos/asf?p=bigtop.git;a=blob;f=bigtop-deploy/puppet/README.md;h=d92e1fc331f18e1e6561a76775813d350fbb88d4;hb=HEAD
