For work I was asked to run cronjobs in a Docker container. These are jobs that will be run on a server related to work. The actual cron script I am currently running in this container consists of just ” filler” jobs that are meant to be silly, and are just being used to debug the process.

I am running a docker container using a Ubuntu image. Here is the Docker file:

FROM ubuntu:latest

# Install necessary software
RUN apt-get update && 
    apt-get -y install cron

# Add crontab file
COPY crontab /etc/cron.d/crontab

# make the directory
#RUN mkdir etc/cron.d

# create a crontab file in a new location 
RUN touch etc/cron.d/crontab


# Give execution rights on the cron job
RUN chmod 644 /etc/cron.d/crontab

# Create the log file to be able to run tail
RUN touch /var/log/cron.log

# Start the cron service
CMD cron && tail -f /var/log/cron.log`

Here is the docker compose ( I changed the username and pword just for privacy, it’s something more specific in the actual file):

version: "3"

services:
  ubuntu_cron_sql_container:
    build:
      context: .
      dockerfile: docker/Dockerfile
    image: ubuntu:latest
    container_name: ubuntu_cron_sql_container
    environment:
      MYSQL_USER: 'username'
      MYSQL_PASSWORD: 'password'
      MYSQL_ROOT_PASSWORD: 'password'
    volumes:
      - ./mysql_server/init.d:/docker-entrypoint-initdb.d
    ports:
      - 3306:3306
    # expose:
    #   - '3306'
    restart: on-failure

(I am planning on adding MySQL to the container next, that’s why it’s in the docker compose, but not in the Dockerfile.)

I am starting the container using the following Bash scripts (these were actually provided to me by someone else at work, I modified them on my own).

I run “build.sh” first:

#!/bin/sh
docker container rm ubuntu_cron_sql_container
docker-compose -f docker-compose.yml build --force-rm

Followed by “run-db.sh”

#!/bin/sh
#docker container rm ubuntu_cron_sql_container

docker-compose -f docker-compose.yml up ubuntu_cron_sql_container

When I access the running container using the Bash shell, and run

ps -ef | grep cron | grep -v gre

I get:

root         1     0  0 17:04 ?        00:00:00 /bin/sh -c cron && tail -f /var/log/cron.log
root         8     1  0 17:04 ?        00:00:00 cron
root         9     1  0 17:04 ?        00:00:00 tail -f /var/log/cron.log

When I run the following at the CL inside the container:

/sbin/service cron start

I get the following output:

* Starting periodic command scheduler cron    

But the log file ( which is indeed where it should be) is still empty:

cat cron.log

(literally no output)

The ‘crontab’ file itself, and it is indeed in etc/cron.d inside the container

MAILTO= ( my work email is here in the actual file)
*/10 * * * * echo 'ABC'  >> /var/log/cron.log 
*/10 * * * * /Users/myname/cron_docker_practice/kitten.sh
*/10 * * * * echo 'TESTING'
*/10 * * * * echo " Running cronjob in Ubuntu container"
*/10 * * * * echo "Cronjob running in ubuntu container" >> /var/log/cron.log

To try to debug this on my own, I tried the above commands to test if cron was running in the Docker container, I checked that the files copied into the right directories, and I tried a few commands to force cron to run after the Docker container is up and running (from inside the container).