Version: 0.0.2

A simple monitoring tool. It can run both as a daemon or from cron.

Given a list of docker containers verifies that all of them are running.

Alerts if disk usages goes over a configured threshold.

Alerts if the 5 minutes load average is over a configured threshold.

E.g. verify no matches for the string 'ERROR' in all log files the last 5 minutes or that the string 'successful' appeared at least 3 times.

$ docker run --rm -v "$PWD":/go/src/ismonitor -w /go/src/ismonitor golang:1.6 bash -c 'go get && go build -v'

The application assumes the presence of the config.json file in the current directory.

For example run it from cron every five minutes:

*/5 * * * * cd /root/ismonitor && ./ismonitor

When executing from cron ismonitor can either do the error reporting to stdout (which cron will pick up and do what it's been configured to do with it) or sent as email through the configured smtp server.

Start it as a daemon using the -d flag. I.e.

$ ./ismonitor -d

It writes the pid in a file named pid and the log output in a file named log in the current directory.

If the application is started with the -d flag while the daemon process is running nothing will happen. I used this to trigger restarts of the daemon process from cron in case it would quit, using a cron entry like this:

0 * * * * cd /root/ismonitor && ./ismonitor -d

I.e. every hour the ismonitor applicaton will be started unless it already is running.

When executing as a daemon the cron_schedule configuration is required in the config.json file.

Error reporting can be done either by writing to stdout or by sending mail through an SMTP server.

If SMTP configuration is included in the configuration file error reporting will be done by sending email. Otherwise the error reporting it done to standard output. If ismonitor is executed in daemon mode it's standard output will be redirected to a file named log. If there is no SMTP configuration the error reporting will hence be found in the log file.

The MIT License (MIT)