clog is a centralized monitoring tool for crontabs, scheduled events, etc, running on other computers. The client executes a script on a Linux machine, and sends the result to the server, which shows jobs in a web based interface, with start and finish time, duration, status (success or failure), script output, etc. The server can send notification alerts by e-mail when jobs fail.

Server installation

clog server is developed in Python, needs PostgreSQL and is tested only in Linux.

git clone https://github.com/zanardo/clog
cd clog
make dev

Create a new user in PostgreSQL, which will own the database. The user should have login role, and should not have superuser, create database or other privileges. Sample:

createuser -P clog

Create a new database in PostgreSQL, owned by the new user:

createdb -E UTF-8 -l en_US.UTF-8 -O clog -T template0 clog

Import the initial database schema:

psql -U clog clog < schema.sql

Configure clogd:

cp config.yml.example config.yml
$EDITOR config.yml

You can run clogd with make run or start if manually:

export CLOGD_CONF=$(pwd)/config.yml
.venv/bin/waitress-serve --host 0.0.0.0 --port 7890 clogd:app

After the installation, access with a browser, with user admin and password admin, create a new user for you, and delete the admin user.

Client installation

clog client is written in Python 2.7 and does not need any special module more than the Python standard library. You can install from source distribution:

make install-cli

Client usage

Run clog one time so it can create its directories:

clog

Now you can create a script inside $HOME/.clog-scripts. For example, the script name could be test.sh.

To run the script:

clog run test.sh

clog will put the result into a queue inside $HOME/.clog-queue, and this queue should be dispatched to the server:

clog send-queue http://servername:27890/

You can schedule both actions on cron.