statuspage-controller
v2.0.2
Published
Automates actions on a Statuspage.io page based on New Relic Alerts.
Downloads
49
Maintainers
Readme
Status Page Controller
Automates actions on a Statuspage.io page based on New Relic Alerts.
It will change the status of status page components if the component names matches the exact name of a New Relic Alert Policy.
Mapping components
Mapping is done by exact name match with optional namespacing for component groups.
Example:
# New Relic alert policy name
Downloads
# Statuspage.io component name
Downloads
If a New Relic incident is created for the "Downloads" policy, status page controller will update the matching "Downloads" component in Statuspage.io.
Component Groups
Component groups are mapped by namespacing the New Relic alert policy with the following convention:
"group-component"
Example mapping of component groups on status.redhat.com:
# New Relic alert policy names
access.redhat.com-Downloads
access.redhat.com-Knowledgebase
developers.redhat.com-Downloads
developers.redhat.com-Blog
# Statuspage.io component groups
access.redhat.com
L Downloads
L Knowledgebase
developers.redhat.com
L Downloads
L Blog
Tresholds
- yellow = Incident is at least 10 minutes old
- orange = Incident is at least 20 minutes old
- red = incident is 30+ minutes old
NOTE: the above can be configured, see below.
Install
npm install statuspage-controller
Usage
With the default config
var StatuspageController = require('statuspage-controller')
var spc = new StatuspageController();
spc.start();
This usage will use all config defaults and expect the following environment variables to be set in order to work:
- NR_API_KEYS - Your New Relic API key(s), this can be a single key or comma separated list of keys
- SPIO_PAGE_ID - Your Statuspage.io Page ID
- SPIO_API_KEY - Your Statuspage.io API key
Specify a config object:
var StatuspageController = require('statuspage-controller')
var config = {
POLL_INTERVAL: 10000,
PORT: 8080,
NR_API_KEYS: process.env.NR_API_KEYS,
SPIO_PAGE_ID: process.env.SPIO_PAGE_ID,
SPIO_API_KEY: process.env.SPIO_API_KEY,
};
var spc = new StatuspageController(config);
spc.start();
Configuring with custom thresholds:
var config = {
POLL_INTERVAL: 10000,
PORT: 8080,
NR_API_KEYS: process.env.NR_API_KEYS,
SPIO_PAGE_ID: process.env.SPIO_PAGE_ID,
SPIO_API_KEY: process.env.SPIO_API_KEY,
THRESHOLDS: [
{
"duration": 600,
"status": "degraded_performance"
},
{
"duration": 1200,
"status": "partial_outage"
},
{
"duration": 1800,
"status": "major_outage"
}
]
};
var spc = new StatuspageController(config);
spc.start();
Thresholds are in seconds. The above will change the component to degraded_performance after 600 seconds, and so on. You can have just one threshold if you want to change the component status to a single status after a given time. For example, if you wanted to go strait to major outage after 10 minutes then you would do:
THRESHOLDS: [
{
"duration": 600,
"status": "major_outage"
}
]
Debug
Enable extra debug logging to the console. Don't enable this in production.
var StatuspageController = require('statuspage-controller')
var config = {
DEBUG: true,
};
var spc = new StatuspageController(config);
spc.start();
API
Methods
There are a few API methods that check the health of the controller and allow you to set component overrides:
- GET /api/healthcheck.json - Check the health of the controller
- POST /api/overrides.json - Create a temporary override of a given statuspage.io component
- GET /api/overrides.json - List current active overrides
Healthcheck
Request
curl http://localhost:8080/api/healthcheck.json
Response
{"message":"New Relic and statuspage.io connections established.","ok":true}
Overrides
Request
curl -X POST -H 'content-type:application/json' http://localhost:8080/api/overrides.json -d '{"component_name": "Downloads", "seconds":60, "new_status":"major_outage"}'
Response
{"message":"Successfully added override","component_name":"Downloads","seconds":60}
Authentication
Optionally you can enable basic auth for the API. You'll need to configure an htpasswrd file with user(s) created with htpasswd
command.
- Either have htpasswd command installed with Apache, or npm-htpasswd
- Create a user and new htpasswd file:
htpasswd -c /path/to/users.htpasswd myuser
- Point the config at the password file:
HTPASSWD_FILE: '/path/to/users.htpasswd
Example:
var StatuspageController = require('statuspage-controller')
var config = {
HTPASSWD_FILE: '/path/to/users.htpasswd',
};
var spc = new StatuspageController(config);
spc.start();
SSL
Optionally you can configure the API server to use SSL. See node.js https
- Either use an existing key and cert of create self-signed cert using the following method:
openssl req -x509 -nodes -days 365 -newkey rsa:2048 -keyout /path/to/selfsigned.key -out /path/to/selfsigned.crt
- Point the config at your key and cert files using the
tls
object:
Example:
var StatuspageController = require('statuspage-controller')
var config = {
TLS: {
key: "/path/to/selfsigned.key",
cert: "/path/to/selfsigned.crt",
}
};
var spc = new StatuspageController(config);
spc.start();
Full Example
Here is an example with all optional config options set:
var StatuspageController = require('statuspage-controller')
var config = {
POLL_INTERVAL: 10000,
PORT: 8080,
DEBUG: false,
NR_API_KEYS: process.env.NR_API_KEYS,
SPIO_PAGE_ID: process.env.SPIO_PAGE_ID,
SPIO_API_KEY: process.env.SPIO_API_KEY,
HTPASSWD_FILE: '/path/to/users.htpasswd',
TLS: {
key: "/path/to/selfsigned.key",
cert: "/path/to/selfsigned.crt",
}
};
var spc = new StatuspageController(config);
spc.start();
Sync vs Push
Both New Relic and StatusPage.io have ways to automate via push. New Relic alerts can post to a webhook, or send an email, and StatusPage.io components can be updated with a unique email. The problem with this method is that if a message is ever lost, then the state between New Relic and StatusPage.io will get out of sync. Also you might not want to change the status of a component until a minimum threshold of failure has occurred.
By synchronizing both systems with their APIs, the states between the two will always be kept in sync, even if an alert message is never received. It also updates states faster than email. As soon as a violation is created it will be picked up on the next sync. If Status Controller ever goes down, the next time it is started it will sync the statuses to their current state.