Description of problem:
When something causes bind restart to fail during ipa-server-install.
ipa-server-install returns a 0 result code.

Version-Release number of selected component (if applicable):
ipa-server-3.0.0-106.20121030T0103zgit63ae7a2.el6.x86_64
bind-dyndb-ldap-2.1-1.el6.x86_64

How reproducible:
Reproducible when bind restart is failing.
Seems to be happening with build from 10-30-2012

Steps to Reproduce:
1. <change setup to make bind restart fail during ipa-server-install>
2. ipa-server-install  --setup-dns --forwarder=10.14.63.12
--hostname=ipaqavmf.testrelm.com -r TESTRELM.COM -n testrelm.com -p <pw> -P
<pw> -a <pw> -U

Actual results:
Configuring DNS (named)
  [1/9]: adding DNS container
  [2/9]: setting up our zone
  [3/9]: setting up reverse zone
  [4/9]: setting up our own record
  [5/9]: setting up kerberos principal
  [6/9]: setting up named.conf
  [7/9]: restarting named
named service failed to start
  [8/9]: configuring named to start on boot
  [9/9]: changing resolv.conf to point to ourselves
Done configuring DNS (named).

Return code: 0

Expected results:
Configuring DNS (named)
  [1/9]: adding DNS container
  [2/9]: setting up our zone
  [3/9]: setting up reverse zone
  [4/9]: setting up our own record
  [5/9]: setting up kerberos principal
  [6/9]: setting up named.conf
  [7/9]: restarting named
named service failed to start
  FAIL - named restart failed
  [8/9]: configuring named to start on boot
  [9/9]: changing resolv.conf to point to ourselves
Failed configuring DNS (named).

Return code: <not 0>

Additional info:
Seems to be a partial regression of
https://bugzilla.redhat.com/show_bug.cgi?id=840381

ipa-server-install always behaved this way. The installation consists of many steps where some errors may pop up. Some of the errors are so critical that installation cannot continue and crashes. In that case we do not return zero return code.

In some cases (like this one), the error is not so critical for installer so we try to continue installing and let user deal with the (minor) issues after the installation.

But I think to make the (minor) issue detecting easier, we could end the section with non-critical issue like as suggested:

  [5/9]: setting up kerberos principal
  [6/9]: setting up named.conf
  [7/9]: restarting named
named service failed to start
  [8/9]: configuring named to start on boot
  [9/9]: changing resolv.conf to point to ourselves
Failed configuring DNS (named).
...
Return code = non-fatal error code, e.g. 2.

Currently, we have just 2 error codes defined:

EXIT STATUS
0 if the (un)installation was successful
1 if an error occurred