ma-zal / pm2-watchdog

Module is continuously checking the health of PM2 NodeJS processes by calling the specified HTTP(s) endpoint. If the HTTP server is unavailable or response is not 2xx, then watchdog will restart the PM2 process.

PM2 module: Webserver Watchdog

Module is continuously checking the health of PM2 NodeJS processes by calling the specified HTTP(s) endpoint. If the HTTP server is unavailable or response is not 2xx, the PM2 process will be restarted.

Installation

pm2 install ma-zal/pm2-watchdog

Basic configuration

# Enable watching for PM2_PROCESS_NAME. Module will continuously checks the availability of URL.
pm2 set pm2-watchdog:url-PM2_PROCESS_NAME URL

Example:

pm2 set pm2-watchdog:url-helloworld http://localhost:3000/status

Optional advanced configuration

# Set the webserver checking for every NUMBER seconds.
# Default is 10 seconds (if no value is specified).
pm2 set pm2-watchdog:checking_interval NUMBER

# Set the app restart when NUMBER webserver checks had failed in a row.
# Default is 3 times (if no value is specified).
pm2 set pm2-watchdog:fails_to_restart NUMBER

# Enable message output level.
# Note: Log files are `~/.pm2/logs/pm2-watchdog-out.log`
#                 and `~/.pm2/logs/pm2-watchdog-error.log`
# Default is `info` (if no value is specified).
# Possible values (levels):
#   `info`  - info messages only (default)
#   `debug` - info + debug messages
#   `trace` - info + debug + trace messages
pm2 set pm2-watchdog:debug LEVEL

# Web request timeout in millis NUMBER .
# Default is 5000 millis.
pm2 set pm2-watchdog:checking_timeout NUMBER

Logging

The processes restarts are logged to error log file (to error output). Therefore if you are tracking errors of your applications, you can redirect the error messages to get complete overview what is happening in your server.

Version Changelog

1.0.1 (17.1.2018)

• Added configurable request timeout