probesys38/agentj_syslogng repository overview

#################### CAUTION ############

This document is not up to date with the new features existing in the V2 of AgentJ

You can see the change ins changelog for the moment

The Document will be update till the end of february

############################################

⁠AgentJ dockerized repository

This is the Docker stack to set up a fully functional AgentJ antispam system at a glance.

⁠Introduction

The AgentJ Docker stack is composed of the following services:

• app: main AgentJ interface (configuration for admins, usage for users)
• db: a MariaDB instance to store mails, user account, DKIM keys, allowed/banned senders, Amavis scores …
• smtp: a postfix instance that will receive the incoming e-mails and forward them to the amavis container (amavis/clamav/spamassassin service)
• outsmtp: a postfix instance that will handle outgoing e-mails, sent by local users (via their original smtp server) and pass them to outamavis
• relay: an other Postfix instance, needed to avoid loops when forwarding the released or white-listed e-mails to their recipients(s)
• amavis: a container running Amavis/Spamassassin and ClamAV services which check incoming mail
• outamavis: same as amavis but used for outgoing e-mails sent by local users
• opendkim: verify incoming mail DKIM signature for incoming mail, and append signature for outgoing mail
• policyd-rate-limit: rate limiting service used by outsmtp, get policies from db
• wip logspout + syslogng: a Syslog-NG instance that will collect and centralize logs from the other containers
• for tests only smtptest and badrelay: see tests⁠ below

⁠Get the sources

All you have to do is to clone the repository

git clone https://github.com/Probesys/agentj.git

then, cd to the cloned repository to configure a few variables:

cd agentj

⁠Configure

⁠Variables

Variables are defined in the .env.example. This file is just a template, you must copy it to .env and adapt variables to you need

⁠Use

It is not recommended to launch the stack as root. We recommend you to create a dedicated docker user (make sure it belongs to the docker group). After you have set the above variables, you can start the stack with following commands:

docker compose up -d

The Web UI will be available at http://hostname:8090⁠. The default login is admin and the default password is changeme.

⁠Development

In .env set COMPOSE_FILE=docker-compose.yml:compose.dev.yml, and DB_EXPOSED_PORT.

⁠Tests

Set COMPOSE_FILE=docker-compose.yml:compose.dev.yml:compose.test.yml, then run docker compose exec -u www-data app ./docker/tests/testmail.sh

⁠Details

⁠Volumes

When started, the AgentJ stack will create the following volumes:

• amavis_in/amavis_out : Amavis databases
• applogs : application logs (cron, web server logs …)
• db : MariaDB databases files
• wip logs: log files from all containers, centralized by the syslogng container
• postqueue : the incoming mail queue (for smtp)
• outpostqueue : the outgoing mail queue (for outsmtp)

⁠Communication matrix

⁠Upgrade

Please read the dedicated documentation⁠ as well as releases notes before upgrading.

Generally speaking, the upgrade processes consists in the following:

docker compose down
# Change VERSION variable in your `.env` file
docker compose up -d

⁠About

⁠License

This work is made available under the GNU Affero General Public License v3.0.

⁠Development

AgentJ is a Probesys⁠ project.