Bug 1017449

Description of problem:

While debugging an issue with syncing a pulp child node, we observed that the pulp configuration is representative of earlier versions of pulp.  As a result, we need to update the /etc/pulp/server.conf that gets installed to more accurately reflect the version of pulp being used and the needs of Satellite 6.

The following is the current state of the server.conf used by the installers:

katello-configure: https://github.com/Katello/katello-installer/blob/master/modules/pulp/templates/etc/pulp/server.conf.erb

node-installer: https://github.com/Katello/node-installer/blob/master/modules/pulp/templates/etc/pulp/server.conf.erb

The following is the current state of the server.conf provided by pulp:

https://github.com/pulp/pulp/blob/master/server/etc/pulp/server.conf

We need to compare the values from pulp with the values used by the installer and update accordingly.  For example, the initial issue that raised this BZ was that the *_timeout values have been updated by pulp (e.g. 30s:12h); however, the timeout we use is 10:600.  See additional notes below for input from Jeff on these.  With the current settings a child sync (which uses update_timeout) may timeout after 10 minutes.

Version-Release number of selected component (if applicable):
node-installer-0.0.15-2.el6sat.noarch
katello-configure-1.4.5-9.el6sat.noarch

How reproducible:
always

Steps to Reproduce:
1. install a Satellite server and run katello-configure
2. install a Satellite node and run node-install
3.

Actual results:
/etc/pulp/server.conf is contains old values

Expected results:
/etc/pulp/server.conf contains updated values

Additional info:

<jortel> I'd bump those timeouts using your own judgment.  the 1st number is how long the agent has to start actually running the request.  The 2nd is how long it has to complete it.  I'd make the 2nd number like 3d (3 days) or more and the 1st, like, 36h (36 hours).  The only down side is; the bigger the 1st setting, the longer it will take for the caller to get feedback that the agent is down.