Ansible Configuration Management Database

About

Ansible-cmdb takes the output of Ansible's fact gathering and converts it into a static HTML overview page containing system configuration information.

It supports multiple templates (html, txt_table, csv, json output, markdown) and extending information gathered by Ansible with custom data. For each host it also shows the groups, host variables, custom variables and machine-local facts.

HTML example output.

Installation

Get the package for your distribution from the Releases page (Not required for MacOS X install)

For Debian / Ubuntu systems:

sudo dpkg -i ansible-cmdb*.deb

For Redhat / Centos systems:

sudo yum --disablerepo=* install ansible-cmdb*.rpm

For MacOS X systems:

brew install ansible-cmdb

For Other systems:

tar -vxzf ansible-cmdb*.tar.gz
cd ansible-cmdb*
sudo make install

Installation from Git repository:

git clone git@github.com:fboender/ansible-cmdb.git
cd ansible-cmdb
sudo make install

Usage

Basic

First, generate Ansible output for your hosts:

mkdir out
ansible -m setup --tree out/ all

Next, call ansible-cmdb on the resulting out/ directory to generate the CMDB overview page:

ansible-cmdb out/ > overview.html

The default template is html_fancy, which uses jQuery.

Full usage

Usage: ansible-cmdb [option] <dir> > output.html

Options:
  --version             show program's version number and exit
  -h, --help            show this help message and exit
  -t TEMPLATE, --template=TEMPLATE
                        Template to use. Default is 'html_fancy'
  -i INVENTORY, --inventory=INVENTORY
                        Inventory to read extra info from
  -f, --fact-cache      <dir> contains fact-cache files
  -p PARAMS, --params=PARAMS
                        Params to send to template
  -d, --debug           Show debug output
  -c COLUMNS, --columns=COLUMNS
                        Show only given columns

Inventory scanning

Ansible-cmdb can read your inventory file (hosts, by default), inventory directory or dynamic inventory and extract useful information from it such as:

All the groups a host belongs to.
Host variables. These are optional key/value pairs for each host which can be used in playbooks. They are scanned by ansible-cmdb and get added to a hosts discovered facts under the 'hostvars' section.

Reading the inventory is done using the -i switch to ansible-cmdb. It takes a single parameter: your hosts file, directory containing your hosts files or path to your dynamic inventory script.

For example:

$ ansible-cmdb -i ./hosts out/ > overview.html

If a host_vars directory exists at that location, it will also be read.

The ''html_fancy'' template uses four extra fields:

groups: A list of Ansible groups the host belongs to.
dtap: Whether a host is a development, test, acceptance or production system.
comment: A comment for the host.
ext_id: An external unique identifier for the host.

For example, let's say we have the following hosts file:

[cust.megacorp]
db1.dev.megacorp.com   dtap=dev  comment="Old database server"
db2.dev.megacorp.com   dtap=dev  comment="New database server"
test.megacorp.com      dtap=test 
acc.megacorp.com       dtap=acc  comment="24/7 support"
megacorp.com           dtap=prod comment="Hosting by Foo" ext_id="SRV_10029"

[os.redhat]
megacorp.com
acc.megacorp.com
test.megacorp.com
db2.dev.megacorp.com

[os.debian]
db1.dev.megacorp.com

The host acc.megacorp.com will have groups 'cust.megacorp' and 'os.redhat', will have a comment saying it has 24/7 support and will be marked as a acc server. Megacorp.com host will have an external ID of "SRV_10029", which will be required by for communicating with Foo company about hosting.

See http://docs.ansible.com/intro_inventory.html#host-variables for more information on host variables.

Any variables set for your hosts will become available in the html_fancy template under the "Custom variables" heading.

Templates

ansible-cmdb offers multiple templates. You can choose your template with the -t or --template argument:

ansible-cmdb -t tpl_custom out/ > overview.html

The 'html_fancy' template is the default.

Ansible-cmdb currently provides the following templates out of the box:

html_fancy: A fancy HTML page that uses jQuery and DataTables to give you a searchable, sortable table overview of all hosts with detailed information just a click away.

It takes a parameter local_js which, if set, will load resources from the local disk instead of over the network. To enable it, call ansible-cmdb with:
```
ansible-cmdb -t html_fancy -p local_js=1 out > overview.html
```
It can be easily extended by copying it and modifying the cols definition at the top.
txt_table: A quick text table summary of the available hosts with some minimal information.
json: The json template simply dumps a JSON-encoded representation of the gathered information. This includes all the extra information scanned by ansible-cmdb such as groups, variables, custom information, etc.
csv: The CSV template outputs a CSV file of your hosts.
markdown: The Markdown template generates host information in the Markdown format.

sql: The SQL template generates an .sql file that can be loaded into an SQLite or MySQL database.

  $ ansible-cmdb -t sql -i hosts out > cmdb.sql
  $ echo "CREATE DATABASE ansiblecmdb" | mysql 
  $ mysql ansiblecmdb < cmdb.sql

You can create your own template or extend an existing one by copying it and refering to the full path to the template when using the -t option:

$ ansible-cmdb -t /home/fboender/my_template out/ > my_template.html

Fact caching

Ansible can cache facts from hosts when running playbooks. This is configured in Ansible like:

[defaults]
fact_caching=jsonfile
fact_caching_connection = /path/to/facts/dir

You can use these cached facts as facts directories with ansible-cmdb by specifying the -f (--fact-cache) option:

$ ansible-cmdb -f /path/to/facts/dir > overview.html

Please note that the --fact-cache option will apply to all fact directories you specify. This means you can't mix fact-cache fact directories and normal setup fact directories. Also, if you wish to manually extend facts (see the Extending chapter), you must omit the ansible_facts key and put items in the root of the JSON.

Columns

Some templates, such as txt_table and html_fancy, support columns. If a template supports columns, you can use the --columns / -c command line option to specify which columns to show.

The --columns takes a comma-separated list of columns (no spaces!) which should be shown. The columns must be specified by their id field. For information on what id fields are supported by a template, take a look in the template. Usually it's the column title, but in lowercase and with spaces replaced by underscores.

For example:

$ ansible-cmdb -t txt_table --columns name,os,ip,mem,cpus facts/
Name                    OS             IP             Mem  CPUs
----------------------  -------------  -------------  ---  -  
jib.electricmonk.nl     Linuxmint 17   192.168.0.3    16g  1  
app.uat.local           Debian 6.0.10  192.168.57.1   1g   1  
eek.electricmonk.nl     Ubuntu 14.04   192.168.0.10   3g   1  
db01.prod.local         Debian 6.0.10  192.168.58.1   0g   1  
debian.dev.local        Debian 6.0.10  192.168.56.2   1g   1  
db02.prod.local         Debian 6.0.10  192.168.58.2   0g   1  
centos.dev.local        CentOS 6.6     192.168.56.8   1g   1  
win.dev.local           Windows 2012   10.0.0.3       4g   0  
host5.example.com       Debian 6.0.10  192.168.57.1   1g   1  
db03.prod.local         Debian 6.0.10  192.168.58.3   0g   1  
zoltar.electricmonk.nl  Ubuntu 14.04   194.187.79.11  4g   2

Extending

You can specify multiple directories that need to be scanned for facts. This lets you override, extend and fill in missing information on hosts. You can also use this to create completely new hosts or to add custom facts to your hosts.

Extended facts are basically the same as normal Ansible fact files. When you specify multiple fact directories, Ansible-cmdb scans all of the in order and overlays the facts.

Note that the host must still be present in your hosts file, or it will not generate anything.

If you're using the --fact-cache option, you must omit the ansible_facts key and put items in the root of the JSON. This also means that you can only extend native ansible facts and not information read from the hosts file by ansible-cmdb.

Override / fill in facts

Sometimes Ansible doesn't properly gather certain facts for hosts. For instance, OpenBSD facts don't include the userspace_architecture fact. You can add it manually to a host.

Create a directory for your extended facts:

$ mkdir out_extend

Create a file in it for a host. The file must be named the same as it appears in your hosts file:

$ vi out_extend/openbsd.dev.local
{
  "ansible_facts": {
      "ansible_userspace_architecture": "x86_64"
  }
}

Specify both directories when generating the output:

./ansible-cmdb out/ out_extend/ > overview.html

Your OpenBSD host will now include the 'Userspace Architecture' fact.

Manual hosts

For example, lets say you have 100 linux machines, but only one windows machine. It's not worth setting up ansible on that one windows machine, but you still want it to appear in your overview...

Create a directory for you custom facts:

$ mkdir out_manual

Create a file in it for your windows host:

$ vi out_manual/win.dev.local
{
  "groups": [
  ],
  "ansible_facts": {
    "ansible_all_ipv4_addresses": [
      "10.10.0.2",
      "191.37.104.122"
    ], 
    "ansible_default_ipv4": {
      "address": "191.37.104.122"
    }, 
    "ansible_devices": {
    }, 
    "ansible_distribution": "Windows", 
    "ansible_distribution_major_version": "2008", 
    "ansible_distribution_release": "", 
    "ansible_distribution_version": "2008", 
    "ansible_domain": "win.dev.local", 
    "ansible_fips": false, 
    "ansible_form_factor": "VPS", 
    "ansible_fqdn": "win.dev.local", 
    "ansible_hostname": "win", 
    "ansible_machine": "x86_64", 
    "ansible_nodename": "win.dev.local", 
    "ansible_userspace_architecture": "x86_64", 
    "ansible_userspace_bits": "64", 
    "ansible_virtualization_role": "guest", 
    "ansible_virtualization_type": "xen", 
    "module_setup": true
  }, 
  "changed": false
}

Now you can create the overview including the windows host by specifying two fact directories:

./ansible-cmdb out/ out_manual/ > overview.html

Custom facts

You can add custom facts (not to be confused with 'custom variables') to you hosts. These facts will be displayed in the html_fancy template by default under the 'Custom facts' header.

Let's say you want to add information about installed software to your facts.

Create a directory for you custom facts:

$ mkdir out_custom

Create a file in it for the host where you want to add the custom facts:

$ vi custfact.test.local
{
  "custom_facts": {
    "software": {
      "apache": {
        "version": "2.4",
        "install_src": "backport_deb"
      },
      "mysql-server": {
        "version": "5.5",
        "install_src": "manual_compile"
      },
      "redis": {
        "version": "3.0.7",
        "install_src": "manual_compile"
      }
    }
  }
}

For this to work the facts must be listed under the custom_facts key.

Generate the overview:

./ansible-cmdb out/ out_custom/ > overview.html

The software items will be listed under the "Custom facts" heading.

Infrequently Asked Questions

Solaris machines have no disk information

Ansible currently does not include disk size information for Solaris hosts. As such, we can't include it in the output of Ansible-cmdb. See issue #24 for more information.

Python packaging / Pypi?

Python has some of the most horrendous packaging infrastructure I've ever encountered in 25 years of programming. As such, anything related to Python packaging will not be supported.

Development

Running from the git repo

If you want to run ansible-cmdb directly from the Git repo:

$ cd ansible-cmdb
$ export PYTHONPATH="$(readlink -f lib)"
$ src/ansible-cmdb

Inner workings

Here's a quick introduction on how ansible-cmdb works internally.

The main section in ansible-cmdb reads the commandline params and instantiates an Ansible object.
The Ansible object first reads in all the facts by calling Ansible.parse_fact_dir() for each argument. This includes the user-extended facts.
If hosts file(s) should be parsed (-i option), ansible calls Ansible.parse_hosts_inventory(). This first reads in all found hosts files into one big string, and then it parses it. For this it uses the AnsibleHostParser class.
The AnsibleHostParser class first parses the inventory and then creates a dictionary with all known ansible node names (hosts) as the keys, but with empty values. It then goes through the 'children', 'vars' and normal sections from the inventory and applies the found information to the hosts dictionary.
When AnsibleHostParser is done, the Ansible class takes all the parsed hosts information and updates its own version of the hosts dictionary.
Finally, the output is generated by the main section.

Updating a host in the Ansible object is done using the Ansible.update_host method. This method does a deep-update of a dictionary. This lets ansible-cmdb overlay information from the facts dir, extended / manual facts and hosts inventory files.

Make targets

For building, make is used. Here are some useful targets:

make test: build some tests.
make release: build a release.
make clean: remove build and other artifacts.

Build packages and source-ball

To build Debian, RedHat and source-packages for ansible-cmdb you'll need a Debian based operating system and you'll have to install the following dependencies:

git
make
python-markdown
zip
fakeroot
alien

You can then build the packages with

make release REL_VERSION=$VERSION

where $VERSION is a (arbitrary) version number.

In order to build releases, your repository will have to be completely clean: everything must be commited and there must be no untracked files. If you want to build a test release, you can temporary stash your untracked changes:

git stash -u

Contributions

If you wish to contribute code, please consider the following:

Any form of Python packaging will NOT be supported. Merge requests involving python packages will not be considered. See issue #23.
Thank you for even considering contributing. I'm quite newbie-friendly, so don't hesitate to ask for help!
Code should be reasonably PEP8-like. I'm not too strict on this.
One logical change per merge request.
By putting in a merge request or putting code in comments, you automatically grant me permission to include this code in ansible-cmdb under the license (GPLv3) that ansible-cmdb uses.
Please don't be disappointed or angry if your contributions end up unused. It's not that they aren't appreciated, but I can be somewhat strict when it comes to code quality, feature-creep, etc.

When in doubt, just open a pull-request and post a comment on what you're unclear of, and we'll figure it out.

Licensing and credits

Ansible-cmdb is licensed under the GPLv3:

This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program.  If not, see <http://www.gnu.org/licenses/>.

For the full license, see the LICENSE file.

Ansible-cmdb started as a short Python script, which I blogged about here:

http://www.electricmonk.nl/log/2015/01/21/host-inventory-overview-using-ansibles-facts/

Cris van Pelt then took that and expanded it into a HTML page. Eventually I forked it to Github and made it public, adding features. Many people collaborated to make Ansible-cmdb into what it is today. For a full list, see the annotations in the CHANGELOG.

xiangxiaobaog3/ansible-cmdb