How can I tell if oo-install worked?

Question

I'm trying to install Openshift Origin on CentOS 6.2 using the portable installer, https://install.openshift.com/portable/oo-install-origin.zip, with all services installed on one host (i.e., broker, app node, MongoDB, ActiveMQ). The oo-install-origin utility seems to have worked (details below) – it did not complain of any errors. I rebooted my server when finished.

• Where is the install log file?
• How can I tell if the install was successful?
• What processes should I see running?
  • activemq is running, but I don't see any MongoDB process. I do see a oo-mongo-setup script in /usr/sbin/. Should I run that, or should the install script have run that?

http://openshift.github.io/documentation/oo_install_users_guide.html#installing-grab-some-tea says

The entire installation process can take anywhere from 10 to 45 minutes. During the process, you will see a fairly constant stream of information scrolling by in your command terminal.

Well, I didn't see a constant stream of information scrolling by in my terminal. Should I have seen more (that what is shown below), or are the docs just out of date? I ran top in another terminal and saw lots of yum activity. However, I have no idea if all of the packages were installed correctly.

Install script output:

# ./oo-install-origin
Checking for necessary tools...
...looks good.
Using bundled assets.
Extracting oo-install to temporary directory...
Starting oo-install...
OpenShift Installer (Build 20140131-2031)
----------------------------------------------------------------------

Welcome to OpenShift.

This installer will guide you through a basic system deployment, based
on one of the scenarios below.

Select from the following installation scenarios.
You can also type '?' for Help or 'q' to Quit:
1. Install OpenShift Origin
2. Add a Node to an OpenShift Origin deployment
3. Generate a Puppet Configuration File
Type a selection and press <return>: 1


Here are the details of your current deployment.

Note: ActiveMQ, MongoDB and named will all be installed on the Broker.
For more flexibility, rerun the installer in advanced mode (-a).

DNS Settings
  * App Domain: xxx.yyy.net
  * Register OpenShift components with OpenShift DNS? No

Role Assignments
+--------+---------------+
| Broker | dell710appeng |
| Node   | dell710appeng |
+--------+---------------+

Host Information
+----------------+---------------+
| Host           | dell710appeng |
| Roles          | Broker, Node  |
| SSH Host       | localhost     |
| User           | root          |
| IP Addr        | 10.xx.xx.xx   |
| BIND DNS Addr  | 10.xx.xx.xx   |
| IP Interface   | em1           |
| Install Status | new           |
+----------------+---------------+

Do you want to change the deployment info? (y/n/q/?) n

Here is the subscription configuration that the installer will use for
this deployment.
+-------------------+--------------------------------------------------------------------+
| Setting           | Value                                                              |
+-------------------+--------------------------------------------------------------------+
| type              | yum                                                                |
| repos_base        | https://mirror.openshift.com/pub/origin-server/release/3/fedora-19 |
| jboss_repo_base   | -                                                                  |
| jenkins_repo_base | http://pkg.jenkins-ci.org/redhat                                   |
| os_repo           | -                                                                  |
| os_optional_repo  | -                                                                  |
+-------------------+--------------------------------------------------------------------+

Do you want to make any changes to the subscription info in the
configuration file? (y/n/q/?) n

Do you want to set any temporary subscription settings for this
installation only? (y/n/q/?) n

Preflight check: verifying system and resource availability.

Checking dell710appeng:
* Target host is running non-Fedora, non-RHEL
* Located getenforce
* SELinux is running in enforcing mode
* Located yum
* Located puppet
* Located augtool
* Located dnssec-keygen
* Located htpasswd
* Located scl
* The ruby193 software collection is installed.
* epel repository is present and enabled

Deploying workflow 'origin_deploy'.
Setting up htpasswd for default user account.

Preparing to install OpenShift Origin on the following hosts:
  * localhost: msgserver, dbserver, broker, node
Deploying host 'dell710appeng'

Checking for xxx.yyy.net DNS key(s) on localhost...
...found at /var/named/Kxxx.yyy.net*.key
Command 'systemctl' didn't work; trying older style...
Older style system command succeeded.

Running Puppet deployment

Running: bash -l -c 'scl enable ruby193 "puppet module uninstall -f openshift/openshift_origin"'
Uninstall command failed; this is expected if the puppet module wasn't previously installed.

Running: bash -l -c 'scl enable ruby193 "puppet module install -v 3.0.1 openshift/openshift_origin"'
Command completed.

Running: bash -l -c 'yum clean all'
Command completed.

Running: bash -l -c 'scl enable ruby193 "puppet apply --verbose /tmp/oo_install_configure_dell710appeng.pp"'
#### The script stopped here for about 20 minutes -- top showed lots of yum activity
Command completed.

Running: bash -l -c 'rm /tmp/oo_install_configure_dell710appeng.pp'
Command completed.
OpenShift Origin deployment completed.
You should manually reboot dell710appeng to complete the process.

All tasks completed.
oo-install exited; removing temporary assets.

Answer 1

The question you ask is more how to control a Puppet operation result and request assistance more at the Puppet level than at the OpenShift level.

In such scenarii with a software installed in an unknown state, the best practice is to take the list of installations operations and check step by step if it has been done.

You're lucky in the case of OpenShift: it could also be installed manually, step by step, with a very well documented process: http://openshift.github.io/documentation/oo_deployment_guide_comprehensive.html

If this is a totally new machine installed only for OpenShift and with no other use, I would suggest to restart from scratch, so you control every step of the process.