Support daemon creation and management on Linux.
Documentation is available at the node-linux portal.
This is a standalone module, originally designed for internal use in NGN. However; it is capable of providing the same features for Node.JS scripts independently of NGN.
This module makes it possible to daemonize Node.js scripts natively (using systemv init.d scripts).
To start, install node-linux via:
npm install node-linux
node-linux has a utility to run Node.js scripts as Linux daemons.
To create a service with node-linux, prepare a script like:
var Service = require'node-linux'Service;// Create a new service objectvar svc =name:'Hello World'description: 'The nodejs.org example web server.'script: '/path/to/helloworld.js';// Listen for the "install" event, which indicates the// process is available as a service.svcon'install'svcstart;;svcinstall;
The code above creates a new
Service object, providing a pretty name and description.
script attribute identifies the Node.js script that should run as a service. Upon running
this, the script will be available to the system. By default, node-linux produces systemv init
scripts, meaning the services can be managed by typing
service myapp start or
service myapp stop
service myapp status in some cases).
Service object emits the following events:
- install - Fired when the script is installed as a service.
- alreadyinstalled - Fired if the script is already known to be a service.
- invalidinstallation - Fired if an installation is detected but missing required files.
- uninstall - Fired when an uninstallation is complete.
- start - Fired when the new service is started.
- stop - Fired when the service is stopped.
- error - Fired in some instances when an error occurs.
- doesnotexist - Fired when an attempt to start a non-existent service is detected.
In the example above, the script listens for the
install event. Since this event
is fired when a service installation is complete, it is safe to start the service.
Services created by node-linux are like other services running on Linux.
They can be started/stopped using
service myapp start or
service myapp stop and
logs are available (default is in /var/log).
Sometimes you may want to provide a service with static data, passed in on creation of the service. You can do this by setting environment variables in the service config, as shown below:
var svc =name:'Hello World'description: 'The nodejs.org example web server.'script: '/path/to/helloworld.js'env:name: "HOME"value: processenv"USERPROFILE" // service is now able to access the user who created its' home directory;
You can also supply an array to set multiple environment variables:
var svc =name:'Hello World'description: 'The nodejs.org example web server.'script: '/path/to/helloworld.js'env:name: "HOME"value: processenv"USERPROFILE" // service is now able to access the user who created its' home directoryname: "TEMP"value: pathjoinprocessenv"USERPROFILE""/temp" // use a temp directory in user's home directory;
Uninstalling a previously created service is syntactically similar to installation.
var Service = require'node-linux'Service;// Create a new service objectvar svc =name:'Hello World'script: require'path'join__dirname'helloworld.js';// Listen for the "uninstall" event so we know when it's done.svcon'uninstall'console.log'Uninstall complete.';console.log'The service exists: 'svcexists;;// Uninstall the service.svcuninstall;
The uninstall process only removes process-specific files. It does NOT delete your Node.js script, but it will remove the logs!
Lots of things!
Long Running Processes & Monitoring:
There is no built-in service recovery in most Linux environments, and third party products can be fairly limited or not easily configured from code. Therefore, node-linux creates a wrapper around the Node.js script. This wrapper is responsible for restarting a failed service in an intelligent and configurable manner. For example, if your script crashes due to an unknown error, node-linux will attempt to restart it. By default, this occurs every second. However; if the script has a fatal flaw that makes it crash repeatedly, it adds unnecessary overhead to the system. node-linux handles this by increasing the time interval between restarts and capping the maximum number of restarts.
Smarter Restarts That Won't Pummel Your Server:
Using the default settings, node-linux adds 25% to the wait interval each time it needs to restart
the script. With the default setting (1 second), the first restart attempt occurs after one second.
The second occurs after 1.25 seconds. The third after 1.56 seconds (1.25 increased by 25%) and so on.
Both the initial wait time and the growth rate are configuration options that can be passed to a new
Service. For example:
var svc =name:'Hello World'description: 'The nodejs.org example web server.'script: '/path/to/helloworld.js'wait: 2grow: 5;
In this example, the wait period will start at 2 seconds and increase by 50%. So, the second attempt would be 3 seconds later while the fourth would be 4.5 seconds later.
Don't DOS Yourself!
Repetitive recycling could potentially go on forever with a bad script. To handle these situations, node-linux
supports two kinds of caps. Using
maxRetries will cap the maximum number of restart attempts. By
default, this is unlimited. Setting it to 3 would tell the process to no longer restart a process
after it has failed 3 times. Another option is
maxRestarts, which caps the number of restarts attempted
within 60 seconds. For example, if this is set to 3 (the default) and the process crashes/restarts repeatedly,
node-linux will cease restart attempts after the 3rd cycle in a 60 second window. Both of these
configuration options can be set, just like
Finally, an attribute called
abortOnError can be set to
true if you want your script to not restart
at all when it exits with an error.
node-linux uses the templates to generate init.d scripts for each Node.js script deployed as a
service. This file is created in
/etc/init.d by default. Additionally, a log file is
/var/log/<name> for general output and error logging.
A log source named
myappname.log provides basic logging for the service. It can be used to see
when the entire service starts/stops.
By default, any
console.error or other output will be made available
in one of these two files.
Due to some unforeseen life circumstances, I was not able to add all of the features I'd hoped to add before releasing this. I'll chip away at them over time, but I would be very interested in community contributions in the following areas:
- systemd script generation
- upstart script generation
I have also added a tag in the issues called
feature request to keep a running to-do list.
If you are interested in working on one of these features, please get in touch with me before you start to discuss the feature.
Copyright (c) 2013 Corey Butler
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the 'Software'), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.