sshc - SSH Configuration Management Tool with Ansible Inventory Generation

This tool can help you manage ssh config files with hosts as well as ansible inventory file.

What it does?

1. It creates a host database (JSON document with metadata + hosts).
2. Create SSH config from that host database.
3. Create Ansible inventory from that same host database.

Example of generated SSH config

# Generated At: 2023-01-24 11:35:25.885044
# sshc Version: 0.3.0

# -- <
Host server1
HostName 192.168.0.100
Port 22
User ubuntu
IdentityFile /home/fahad/.ssh/id_rsa
LogLevel INFO
Compression yes
# Comment: Personal Server: ONE
# -- >

# -- <
Host server2
HostName 10.10.0.102
Port 4522
User root
IdentityFile /home/fahad/.ssh/id_rsa
LogLevel DEBUG
Compression no
# Comment: Personal Server: TWO
# -- >

Example of generated Ansible Inventory

{
    "all": {
        "hosts": {
            "server1": {
                "ansible_host": "192.168.0.100",
                "ansible_port": 22,
                "ansible_user": "ubuntu",
                "ansible_ssh_private_key_file": "/home/fahad/.ssh/id_rsa"
            },
            "server2": {
                "ansible_host": "10.10.0.102",
                "ansible_port": 4522,
                "ansible_user": "root",
                "ansible_ssh_private_key_file": "/home/fahad/.ssh/id_rsa"
            }
        },
        "children": {
            "personal": {
                "hosts": {
                    "server1": null,
                    "server2": null
                }
            },
            "home": {
                "hosts": {
                    "server1": null
                }
            },
            "storage": {
                "hosts": {
                    "server2": null
                }
            }
        }
    },
    "others": {
        "generated_at": "2023-01-24 11:35:25.885044",
        "sshc_version": "0.3.0"
    }
}

Why?

Problem it tried to solve

• Working with a bunch of servers gets messy to track those down.
• Managing Ansible Inventory and also SSH config file separate is redundant.

Tried to solve via

• Using a JSON file as a common database of hosts (with creation/update metadata and a SHA-256 integrity checksum).
• Setting name, ports, user, private key, ssh compression, ssh connection log level etc when inserting a host information.
• Set groups, do comment on specific host for host management.
• Well sorted config files.
• Ansible inventory is managed using JSON file (YAML/YML also supported).
• Add host to multiple groups which end up with ansible hosts group.
• Remove and update host entry easily (partial update only changes fields you pass).

Description

Structure

1. Insert host information to a JSON file as a DB.
2. Generate SSH Config file and an Ansible Inventory file.

Technology Stack

1. python
2. json
3. yaml
4. openssh
5. ansible

Dependency

Runtime

• Python 3.7+
• Linux or Windows (paths use the user home directory via pathlib, typically ~/.ssh or %USERPROFILE%\.ssh)

Development

• Poetry

Installation

% pip3 install sshc --upgrade

See CHANGELOG.md for release history and unreleased changes.

Usage

Default files live under your user home .ssh directory (cross-platform):

Artifact	Default path
Host DB	<home>/.ssh/sshc_db.json
SSH config	<home>/.ssh/sshc_ssh_config
Ansible inventory	<home>/.ssh/sshc_ansible_inventory.json

On Linux/macOS that is usually $HOME/.ssh/.... On Windows it is typically %USERPROFILE%\.ssh\....

Step 1: Need the DB to be initiated for the first time

Pattern

usage: sshc init [-h] [--destination DESTINATION] [--dbfile DBFILE]

options:
  -h, --help            show this help message and exit
  --destination DESTINATION
                        Config HOME?
  --dbfile DBFILE       SSHC DB File.

Example

% sshc init

This creates <home>/.ssh/sshc_db.json as a document with meta (created/updated timestamps and sshc versions, update_count, hosts_sha256) and an empty hosts array.

If the DB file already exists as a legacy JSON array ([] or a list of hosts), init upgrades it in place to the document format and adds metadata without deleting hosts.

Step 2: Insert host information to the Database

Pattern

usage: sshc insert [-h] --name NAME --host HOST [--user USER] [--port PORT] [--comment COMMENT] [--loglevel {INFO,DEBUG,ERROR,WARNING}] [--compression {yes,no}]
                   [--groups GROUPS [GROUPS ...]] [--identityfile IDENTITYFILE] [--destination DESTINATION] [--dbfile DBFILE]

options:
  -h, --help            show this help message and exit
  --name NAME           Server Name?
  --host HOST           SSH Host?
  --user USER           SSH User? (default: root)
  --port PORT           SSH Port? (default: 22)
  --comment COMMENT     Host comment.
  --loglevel {INFO,DEBUG,ERROR,WARNING}
                        SSH Log Level.
  --compression {yes,no}
                        SSH Connection Compression.
  --groups GROUPS [GROUPS ...]
                        Which groups to include? (space-separated)
  --identityfile IDENTITYFILE
                        SSH identity file location (default: <home>/.ssh/id_rsa)
  --destination DESTINATION
                        Config HOME?
  --dbfile DBFILE       SSHC DB File.

Example

% sshc insert --name Google --host 8.8.8.8 --port 22 --user groot --identityfile /home/fahad/fahad.pem --comment "This is the server where you are not authorized to have access." --groups google fun

If a host name already exists, insert is skipped and a message is printed.

Step 3: Generate ssh config and as well as ansible inventory file

Pattern

usage: sshc generate [-h] [--configfile CONFIGFILE] [--inventoryfile INVENTORYFILE] [--destination DESTINATION] [--dbfile DBFILE] [--filetype {json,yaml,yml}]

options:
  -h, --help            show this help message and exit
  --configfile CONFIGFILE
                        SSH Config File.
  --inventoryfile INVENTORYFILE
                        Ansible Inventory File.
  --destination DESTINATION
                        Config HOME?
  --dbfile DBFILE       SSHC DB File.
  --filetype {json,yaml,yml}
                        Preferred file type for Ansible inventory. Default is json and you can choose yaml too.

Example

% sshc generate

This command will read all the entries in the DB and generate

1. SSH config file in your preferred directory or the default (<home>/.ssh/sshc_ssh_config).
2. Ansible inventory file in your preferred directory or the default (<home>/.ssh/sshc_ansible_inventory.json).

If you stick with the default directory you will find the generated files in:

1. Default Directory: <home>/.ssh
2. Generated Ansible Inventory: <home>/.ssh/sshc_ansible_inventory.json
3. Generated SSH Config: <home>/.ssh/sshc_ssh_config

You can use these configs like below.

For SSH,

% ssh -F $HOME/.ssh/sshc_ssh_config

For Ansible,

% ansible -i $HOME/.ssh/sshc_ansible_inventory.json all --list-host

On Windows (PowerShell), use the same idea with your profile path, for example:

ssh -F $env:USERPROFILE\.ssh\sshc_ssh_config

Note: If you choose default SSH config file location and ansible host file location, sshc will replace the file. Be careful.

Recommended Way of Generating Configurations

• There are two terms to keep in mind.
  • SSH default
  • sshc default
• Use sshc default paths which is different from SSH and Ansible default config.
• Use those newly created files(which should be separate than default one) either passing -F for SSH and -i for Ansible.

Others

Help message of the tool

% sshc --help

usage: sshc [-h] [--version] {init,insert,delete,update,read,generate,status} ...

SSH Config and Ansible Inventory Generator !

options:
  -h, --help            show this help message and exit
  --version             show program's version number and exit

subcommands:
  The main command of this CLI tool.

  {init,insert,delete,update,read,generate,status}
                        The main commands have their own arguments.
    init                Initiate Host DB !
    insert              Insert host information !
    delete              Delete host information !
    update              Update host information !
    read (list)         Read / list host database !
    generate            Generate necessary config files !
    status              Check DB integrity and generated file sync !

Delete Inserted Data

% sshc delete --hostname <HOSTNAME>

Optional path overrides: --destination, --dbfile.

Update Inserted Data

Only fields you pass are changed; omitted fields stay as they are in the DB.

usage: sshc update [-h] --name NAME [--host HOST] [--user USER] [--port PORT] [--comment COMMENT]
                   [--loglevel {INFO,DEBUG,ERROR,WARNING}] [--compression {yes,no}] [--groups GROUPS [GROUPS ...]]
                   [--identityfile IDENTITYFILE] [--destination DESTINATION] [--dbfile DBFILE]

options:
  -h, --help            show this help message and exit
  --name NAME           Server Name?
  --host HOST           SSH Host?
  --user USER           SSH User?
  --port PORT           SSH Port?
  --comment COMMENT     Host comment (omit to leave unchanged).
  --loglevel {INFO,DEBUG,ERROR,WARNING}
                        SSH Log Level (omit to leave unchanged).
  --compression {yes,no}
                        SSH Connection Compression (omit to leave unchanged).
  --groups GROUPS [GROUPS ...]
                        Which group to include? (omit to leave unchanged)
  --identityfile IDENTITYFILE
                        SSH identity file (omit to leave unchanged)
  --destination DESTINATION
                        Config HOME?
  --dbfile DBFILE       SSHC DB File.

Example

% sshc update --name google --port 2222

If the host name is not in the DB, update will insert instead (then --host and --port are required).

Read / list DB Data

list is a synonym for read (same options and output).

% sshc read
% sshc list

You can pass verbose too (prints DB metadata and hosts)

% sshc read --verbose yes
% sshc list --verbose yes

Status / health check

Check DB integrity, emptiness, identity-file paths, and whether SSH config / Ansible inventory match the DB (or need sshc generate).

% sshc status
% sshc status --json
% sshc status --dbfile PATH --configfile PATH --inventoryfile PATH

• Exit code 0 when there are no error-level issues (warnings alone are OK).
• Exit code 1 when integrity fails, generated files are missing/out of sync while hosts exist, or the DB is unusable.
• Missing generated files with an empty DB are warnings; with hosts present they are errors and set needs_regeneration.

Known issues or Limitations

• Tested on Ubuntu 22.04 and Windows (path handling uses pathlib.Path.home(), typically %USERPROFILE%\.ssh on Windows).
• OpenSSH client / Ansible availability on Windows depends on your local install; sshc itself generates the config and inventory files portably.

Getting help

If you have questions, concerns, bug reports and others, please file an issue in this repository's Issue Tracker.

Getting involved

If you want to contribute to this tool, feel free to fork the repo and create Pull request with your changes. Keep in mind to

• include better comment to understand.
• create PR to development branch.
• check the Context/ directory for maintainer-oriented architecture notes, gotchas, and todo tracking.

---

Author

• Fahad Ahammed - DevOps Enthusiast - Dhaka, Bangladesh