Puppet fails but succeeds on the 2nd attempt

[Rajul Vora]

I have installed two modules under $confdir/modules:

erwbgy's system module (which depends on augeasproviders) and augeasproviders

I have a hiera entry:

system::sshd:

  Banner:

    value: /etc/issue.net

When I do the first "puppet agent --test" run, it always fails with following error:

Error: Could not retrieve catalog from remote server: Error 400 on SERVER: Could not autoload puppet/type/sshd_config: Could not autoload puppet/provider/sshd_config/augeas: no such file to load -- augeasproviders/provider on node ip-10-28-37-186.ec2.internal

Warning: Not using cache on failed catalog

Error: Could not retrieve catalog; skipping run

However, I run "puppet agent --test" again and it always works.

This pattern repeats anytime there was a puppet run failure for any reason (like syntax error in one of the unrelated manifests files).

I do have "pluginsync=true" in agent's /etc/puppet/puppet.conf file.

Not sure what I need to do fix this. Any pointers appreciated.

Rajul