github.com/tiredofit/docker-alpine

About

Dockerfile to build an alpine linux container image.

• Currently tracking 3.5, 3.6, 3.7, 3.8, 3.9, 3.10, 3.11, 3.12, 3.13, 3.14, 3.15, 3.16, 3.17, 3.18 , 3.19 and edge.
• s6 overlay enabled for PID 1 init capabilities.
• zabbix-agent (Classic and Modern) for individual container monitoring.
• Scheduling via cron with other helpful tools (bash, curl, less, logrotate, nano, vi) for easier management.
• Messaging ability via MSMTP enabled to send mail from container to external SMTP server.
• Firewall included with capabilities of monitoring logs to block remote hosts via Fail2ban
• Logshipping capabilities to remote log analysis servers via Fluent-Bit
• Ability to update User ID and Group ID permissions dynamically.

Maintainer

• Dave Conroy

Table of Contents

• About
• Maintainer
• Table of Contents
• Prerequisites and Assumptions
• Installation
  • Build from Source
  • Prebuilt Images
    • Multi Architecture
• Configuration
  • Quick Start
  • Persistent Storage
  • Environment Variables
    • Container Options
    • Scheduling Options
      • Cron Options
    • Messaging Options
      • MSMTP Options
    • Monitoring Options
      • Zabbix Options
    • Logging Options
      • Log Shipping Parsing
      • Fluent-Bit Options
    • Firewall Options|
      • IPTables Options
      • Fail2Ban Options
    • Permissions
    • Process Watchdog
  • Networking
• Developing / Overriding
• Debug Mode
• Maintenance
  • Shell Access
• Support
  • Usage
  • Bugfixes
  • Feature Requests
  • Updates
• License

Prerequisites and Assumptions

No prerequisites required

Installation

Build from Source

Clone this repository and build the image with docker build <arguments> (imagename) .

Prebuilt Images

Builds of the image are available on Docker Hub

docker pull docker.io/tiredofdit/alpine:(imagetag)

Builds of the image are also available on the Github Container Registry

docker pull ghcr.io/tiredofit/docker-alpine:(imagetag)

The following image tags are available along with their tagged release based on what's written in the Changelog:

Multi Architecture

Images are built primarily for amd64 architecture, and may also include builds for arm/v7, arm64 and others. These variants are all unsupported. Consider sponsoring my work so that I can work with various hardware. To see if this image supports multiple architecures, type docker manifest (image):(tag)

Configuration

Quick Start

Utilize this image as a base for further builds. Please visit the s6 overlay repository for instructions on how to enable the S6 init system when using this base or look at some of my other images which use this as a base.

Persistent Storage

The following directories are used for configuration and can be mapped for persistent storage.

Environment Variables

Below is the complete list of available options that can be used to customize your installation. Variables showing an 'x' under the _FILE column can be used for storing the information inside of a file, useful for secrets.

Container Options

Scheduling Options

The image has capability of executing tasks at differnt times of the day. It follows the cron syntax. Presently this image only supports the busybox cron however can be extended to other scheduling backends with little effort.

Cron Options

There are two ways to add jobs to be triggered via cron. One is to drop files into /assets/cron/ which will be parsed upon container startup, or to set environment variables.

Example: CRON_HELLO="* * * * * echo 'hello' > /tmp/hello.log

Since you can't really disable environment variables in Docker if they are baked into parent Docker images, you can override a baked in Cron command with your own values, or to disable it entirely set the value to FALSE eg CRON_HELLO=FALSE.

Messaging Options

If you wish to send mail, set CONTAINER_ENABLE_MESSAGING=TRUE and configure the following environment variables. Presently we only support one backend, but more can be added with little effort.

MSMTP Options

See the MSMTP Configuration Options for further information on options to configure MSMTP.

See The Official Zabbix Agent Documentation for information about the following Zabbix values.

Monitoring Options

This image includes the capability of using agents inside the image to monitor metrics from applications. Presently at this time it only supports Zabbix as a monitoring platform, however is extendable to other platforms with little effort.

Zabbix Options

This image comes with Zabbix Agent 1 (Classic or C compiled) and Zabbix Agent 2 (Modern, or Go compiled). See which variables work for each version and make your agent choice. Drop files in /etc/zabbix/zabbix_agentd.conf.d to setup your metrics. The environment variables below only affect the system end of the configuration. If you wish to use your own system configuration without these variables, change ZABBIX_SETUP_TYPE to MANUAL

This image supports autoregistering configuration as an Active Agent to a Zabbix Proxy or a Server. It looks in /etc/zabbix_agent.conf.d/*.conf for the string # Autoregister= and takes these values and adds it to the HostMetadata configuration entry automatically wrapped around : eg :application: . Use it by creating an Auto register rule and search for that string. You can find server templates in this repository in the [zabbix_templates](zabbix_templates/) directory.

Logging Options

This is work in progress for a larger logging solution. Presently there is functionality to rotate logs on a daily basis, however as this section matures there will be the capability to also ship the logs to an external data warehouse like Loki, or Elastic Search. At present Log shipping is only supported by fluent-bit and x86_64 only.

Log Shipping Parsing

You can set an environment variable to start shipping a log without any other configuration. Create a variable starting with LOGSHIP_<name> with the value of the location of the log files. You can also use this to null an existing configuration by setting the value of FALSE.

Example: LOGSHIP_NGINX=/var/log/nginx/*.log will create and tag all log files from that directory as to be coming from CONTAINER_NAME and from nginx. Note, it does not allow for custom parsing, so will simply parse the log entry as is.

If LOGSHIPPING_AUTO_CONFIG_LOGROTATE set to true, you can define what parser the configuration file should use. Make sure you have the appropriate .conf parsers in /etc/fluent-bit/parsers.d/ Create a line in the logrotate.d/<file> that looks like # logship: <parser>. Multiple parsers can be added by seperating via commas. Alternatively, if you wanted to skip a certain logfile from being parsed by the log shipper, use the value "SKIP".

Fluent-Bit Options

Drop files in /etc/fluent-bit/conf.d to setup your inputs and outputs. The environment variables below only affect the system end of the configuration. If you wish to use your own system configuration without these variables, change FLUENTBIT_SETUP_TYPE to MANUAL. Container will attempt to automatically create configuration to send to a destination, or can also be set to act as a receiver from other fluent-bit hosts and forward data to a remote log analysis service.

Firewall Options|

Included when proper capabilities are set on image is the capability to set up detailed block / allow rules via a firewall on container start. Presently only iptables is supported. You must use run your containers with the following capabilities added: NET_ADMIN, NET_RAW

One can use the FIREWALL_RULE_XX environment variables to pass rules to the firewall. In this example I am going to block someone from being able to access a port except if from a specific IP address:

FIREWALL_RULE_00=-I INPUT -p tcp -m tcp -s 101.69.69.101 --dport 389 -j ACCEPT
FIREWALL_RULE_01=-I INPUT -p tcp -m tcp -s 0.0.0.0/0 --dport 389 -j DROP

IPTables Options

Instead of relying on environment variables one can put a iptables-restore compatible ruleset below and it will be imported on container start.

Fail2Ban Options

The container also has the capability should CONTAINER_ENABLE_FIREWALL=TRUE be enabled to launch Fail2ban, a process which watches logs for patterns and then blocks the remote host from connecting for a period of time. Drop your custom jail configs as *.conf files in /etc/fail2ban/jail.d/ and filters in /etc/fail2ban/filter.d for them to be parsed at startup. Note the startup delay environment variable to avoid the process failing if no log files exist from a fresh install.

Permissions

If you wish to change the internal id for users and groups you can set environment variables to do so. e.g. If you add USER_NGINX=1000 it will reset the containers nginx user id from 82 to 1000 -

If you enable DEBUG_PERMISSIONS=TRUE all the users and groups have been modified in accordance with environment variables will be displayed in output.

Hint, also change the Group ID to your local development users UID & GID and avoid Docker permission issues when developing.

Process Watchdog

This is experimental functionality to call an external script before a process is executed.

Sample use cases:

• Alert slack channel when process has executed more than once
• Disable process from executing further if restarted 50 times
• Write to an additional log file..
• Change a file to display "Under Maintenance" on a webserver if this process isn't supposed to be run more than 1 time.

It will pass 5 arguments to a bash script titled the same name as the executing script or if not found, use the default CONTAINER_PROCESS_HELPER_SCRIPT below. Drop your files into the CONTAINER_PROCESS_HELPER_PATH.

For example, if 04-scheduling was starting, it would look for $CONTAINER_PROCESS_HELPER_PATH/04-scheduling and if found execute it while passing the following arguments: DATE,TIME,SCRIPT_NAME,TIMES EXECUTED,HOSTNAME

e.g: 2021-07-01 23:01:04 04-scheduling 2 container

Use the values in your own bash script using the $1 $2 $3 $4 $5 syntax. Change time and date and settings with these environment variables

Networking

The following ports are exposed.

Developing / Overriding

This base image has been used over a hundred times to successfully build secondary images. My methodology is admittedly straying from the "one process per container" rule, however this methodology allows me to put together images at a rapid pace, and if more complex scalability is required, the work is split into their own individual images. Since you are reading this here's a crash course at how the image works: (WIP):

See /assets/functions/00-container for more detailed documentation for the various commands and functions/shortcuts

• Put defaults in /assets/defaults/(script name)

• Put functions in /assets/functions/(script name)

• Put Initialization script in /etc/cont-init.d/(script name)

  Put at the top:

#!/command/with-contenv bash          # Pull in Container Environment Variables from Dockerfile/Docker Runtime
source /assets/functions/00-container # Pull in all custom container functions from this image
prepare_service single                # Read functions and defaults only from files matching this script filename - see detailed docs for more
PROCESS_NAME="process"                # set the prefix for any logging

.. your scripting ..
print_info "This an INFO log"
print_warn "This a WARN log"
print_error "This is a ERROR log"

liftoff                               # this writes to the state files at /tmp/.container/ to prove the script executed properly see CONTAINER_SKIP_SANITY_CHECK

• Put Services script in /etc/services.available/(script name)

  Put at the top:

#!/command/with-contenv bash          # Pull in Container Environment Variables from Dockerfile/Docker Runtime
source /assets/functions/00-container # Pull in all custom container functions from this image
prepare_service defaults single       # Read defaults only from files matching this script filename - see detailed docs for more
PROCESS_NAME="process"                # set the prefix for any logging

check_container_initialized           # Check to make sure that the container properly initialized before proceeding
check_service_initialized init        # Check to see if the cont-init.d/scriptname executed correctly, otherwise wait until it has done
liftoff                               # Prove script was able to execute properly

print_start "Starting processname"    # Show STARTING log prefix, and also show if enabled a counter, and execute process watchdog script
fakeprocess (args)                    # whatever your process you want to start is

Debug Mode

When using this as a base image, create statements in your startup scripts to check for existence of DEBUG_MODE=TRUE and set various parameters in your applications to output more detail, enable debugging modes, and so on. In this base image it does the following:

• Sets zabbix-agent to output logs in verbosity
• Shows all script output (equivalent to set -x)

Maintenance

Shell Access

For debugging and maintenance purposes you may want access the containers shell.

bash docker exec -it (whatever your container name is) bash

Support

These images were built to serve a specific need in a production environment and gradually have had more functionality added based on requests from the community.

Usage

• The Discussions board is a great place for working with the community on tips and tricks of using this image.
• Sponsor me for personalized support

Bugfixes

• Please, submit a Bug Report if something isn't working as expected. I'll do my best to issue a fix in short order.

Feature Requests

• Feel free to submit a feature request, however there is no guarantee that it will be added, or at what timeline.
• Sponsor me regarding development of features.

Updates

• Best effort to track upstream changes, More priority if I am actively using the image in a production environment.
• Sponsor me for up to date releases.

License

MIT. See LICENSE for more details.