Skip to content
This repository has been archived by the owner on Sep 9, 2024. It is now read-only.

Setup AWS Monitoring with Terraform and NodePing

Notifications You must be signed in to change notification settings

softkraftco/terraform-nodeping

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

98 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Terraform NodePing Provider

This is NodePing provider for Terraform. It can be used to manage NodePing resources like Checks or Contacts.

Quick start

First of all, build and install the provider into your local Terraform installation:

git clone https://github.com/softkraftco/terraform-nodeping.git
cd terraform-nodeping
make install

Next create new Terraform project:

mkdir mynodeping
cd mynodeping
touch mynodeping.tf 

As a next step add NodePing provider to required_providers section of mynodeping.tf file:

terraform {
  required_providers {
    nodeping = {
      version = "0.0.1"
      source  = "softkraft.co/terraform/nodeping"
    }
  }  
}

provider "nodeping" {
    token = "00AAAAAA-A0A0-AAAA-0000-A0AA0A000AAA"
}

Then add NodePing check resource definition to mynodeping.tf file:

resource "nodeping_check" "mycheck"{
	label = "mycheck"
	type = "HTTP"
	target = "http://example.eu/"
}

And finally execute Terraform!

terraform apply

Usage

To authenticate to NodePing API, one needs to use an API token. It can be passed to the provider plugin in two ways. The preferred approach is to define a token in a provider block in your Terraform file:

provider "nodeping" {
  token = "00AAAAAA-A0A0-AAAA-0000-A0AA0A000AAA"
}

As an alternative, you can specify token using an environment variable:

export NODEPING_API_TOKEN=00AAAAAA-A0A0-AAAA-0000-A0AA0A000AAA
terraform apply

Resources

This provider supports the following type of resources:

  • Accounts
  • Check
  • Contact
  • Contactgroups
  • Schedule

Accounts

It is an example declaration of account resource. Note that Nodeping in their documentation keeps using Accountsand for the definition type uses customer that is why we use accounts for documentation purposes, but nodeping_customer for the definition.

resource "nodeping_customer" "new_subaccount"{
	name = "NewSubaccount"
	contactname = "Contact Name"
	email = "[email protected]"
	timezone = "1.0"
	location = "nam"
	emailme = false
}

Please follow the official documentation API documentation.

  • name name of the subaccount, used as a label.
  • contactname represents a name of the primary contact for the subaccount, used as a label.
  • email primary email address for the subaccount.
  • timezone main time zone for the subaccount.
  • location location for the subaccount, available values: nam, eur, eao, wlw.
  • emailme to opt-in the subaccount for NodePing service and features email notifications.
  • status subaccount status, available values: "Active" or "Suspend".

Check

This is a simple example declaration of check resource:

resource "nodeping_check" "some_check"{
	label = "SomeCheck"
	type = "HTTP"
	target = "http://example.eu/"
	enabled = "inactive"
}

The implementation generally follows API documentation, with some differences.

  • customer_id a customer id of the subaccount the check belongs to.
  • type string check type, one of: AGENT, AUDIO, CLUSTER, DOHDOT, DNS, FTP, HTTP, HTTPCONTENT, HTTPPARSE, HTTPADV, IMAP4, MYSQL, NTP, PING, POP3, PORT, PUSH, RBL, RDP, SIP, SMTP, SNMP, SPEC10DNS, SPEC10RDDS, SSH, SSL, WEBSOCKET, WHOIS
  • target check target. Note that for HTTP, HTTPCONTENT, HTTPPARSE, and HTTPADV check, this must begin with "http://" or "https://".
  • label gives this check a label. If this is absent, the target will be used as the label.
  • interval how often this check runs in minutes. Can be any integer starting at 1 for one-minute checks. Once a day is 1440. Defaults to 15.
  • enabled sets to 'active' to enable this check to run, set to 'inactive' to disable.

enabled differs from API in that it only accepts strings: 'active', 'inactive', and it defaults to 'active'.

  • public sets to true to enable public reports for this check.

public differs from API in that it uses boolean values instead of strings.

  • runlocations an array of geographical regions where probe servers are located. This can be one or more of the following: 'nam' for North America, 'eur' for Europe, 'eao' for East Asia/Oceania, or 'wlw' for worldwide. If omitted, the account's default location setting is used. To run checks on an AGENT, set the location to the ID of the agent check you want to run on.

Running checks on an AGENT was never tested

runlocations differs from API in that it only allows array as an input value.

  • homeloc sets the preferred probe location, home location, for this check. The default is 'false' and will run the check on a random probe in the selected region (see runlocations) or the account default region if no region is specified on the check. The probe two-letter indicators are listed in our FAQ (example: 'ca' would run a check from our California probe). Set this value to 'roam' to have the check change probe location on each interval.

homeloc differs from API in that it will only accept strings. If you need to run the check on a random probe, set this parameter to "false" (string).

  • threshold the timeout for this check in seconds, defaults to 5 for a five-second timeout. Can be any integer starting at 1. For CLUSTER checks, this indicates how many checks listed in the 'data' element must be passing in order for this check to pass.
  • sens number of rechecks before this check is considered 'down' or 'up'. Defaults to 2 (2 rechecks means it will be checked by 3 different servers). Rechecks are run immediately, not based on the interval, and happen when a status change occurs.
  • notifications array of objects containing the contact id, delay, and scheduling for notifications. The IDs can be obtained by listing the relevant contacts. See more robust example below for usage.

Unlike in the case of API, there is no need to set the contact key value to "None", in order to remove an address from a check's notifications. Simply removing the address item is enough.

  • dep - optional string - the id of the check used for the notification dependency. If the check is set to be failing, no notifications will be sent. For example, set this to the check id of a PING check on an edge router for all services that depend on that router. It helps reduce the number of alerts you receive when core networks or services go offline. To remove this functionality, set this to false.
  • description - optional string - you can put arbitrary text, JSON, XML, etc. The size limit is 1000 characters.

The following are only relevant for certain types:

  • checktoken read-only field on AGENT and PUSH checks - can request server-side re-generation by setting this field to 'reset'.
  • clientcert string to specify the ID of a client certificate/key to be used in the DOHDOT check.
  • contentstring string for DOHDOT, DNS, HTTPCONTENT, HTTPADV, FTP, SSH, WEBSOCKET, WHOIS type checks - the string to match the response against.
  • dohdot string is used to specify DoH or DoT in the DOHDOT check. Valid value is either 'doh' or 'dot' - defaults to 'doh'.
  • dnstype string for DNS and DOHDOT checks to indicate the type of DNS entry to query - String set to one of: 'ANY', 'A', 'AAAA', 'CNAME', 'MX, 'NS, 'PTR', 'SOA', 'SRV', 'TXT'.
  • dnstoresolve optional string for DNS and DOHDOT checks - The FQDN of the DNS query
  • dnsrd boolean for DNS RD (Recursion Desired) bit - defaults to true. If you're using CloudFlare DNS servers, set this to false.
  • transport string for DNS check transport protocol - defaults to 'udp' but 'tcp' is also supported.
  • follow boolean is used for HTTP, HTTPCONTENT, and HTTPADV checks. If true, the check will follow up to four redirects. The HTTPADV check only supports follow for GET requests.
  • email string used for IMAP4 and SMTP checks.
  • port positive integer for DNS, FTP, NTP, PORT, SSH type checks - used for check types that support port fields separate from the target address. HTTP and HTTPCONTENT will ignore this field as the port must be set in the target in standard URL format. This field is required by PORT and NTP checks.
  • username string for FTP, IMAP4, POP3, SMTP, and SSH type checks - HTTP and HTTPCONTENT will ignore this field as the username must be set in the target in standard URL format.
  • password string for FTP, IMAP4, POP3, SMTP, and SSH type checks - Note that this is going to be passed back and forth in the data, so you should always be sure that credentials used for checks are very limited in their access level. See our Terms of Service. HTTP and HTTPCONTENT will ignore this field as the password must be set in the target in standard URL format.
  • secure string to specify whether the IMAP4, SMTP, and POP3 checks should use TLS for the check. Can be set to "false" or "ssl".
  • verify string to set whether or not to verify the SSL certificate (SMTP, IMAP4, POP3, DOHDOT check types) or DNSSEC (DNS check type only). Can be "true" or "false".
  • ignore string for the RBL check type, specifies RBL lists to ignore. Multiple lists can be added by including them in the string, separated by commas.
  • invert string for FTP, HTTPCONTENT, HTTPADV, NTP, PORT, SSH type checks - used for 'Does not contain' functionality in checks. Default is 'false' - Set to 'true' to invert the content type match.
  • warningdays positive integer for SSL, WHOIS, POP3, IMAP4, and SMTP checks - number of days before certificate (or domain for WHOIS) expiration to fail the check and send a notification.
  • fields is used for fields to parse from the HTTPADV, HTTPPARSE, and SNMP response. Each block should include elements: name, min, and max.
  • postdata string that can be used in the HTTPADV check as an alternative to the data object. postdata should be a single string to post.
  • data, receiveheaders, sendheaders are optional maps used by HTTPADV ('data' can also be used for CLUSTER checks - see blow - 'sendheaders' can be used for DOHDOT and HTTPPARSE). They are formatted as key:value pairs.
  • edns object is used to send EDNS(0) OPT psudo-records in a DNS query in the DOHDOT check type. Must be formatted as key:value pairs
  • method string used by the HTTPADV and DOHDOT checks to specify the HTTP method. For the HTTPADV check, the value can be one of GET, POST, PUT, HEAD, TRACE, or CONNECT. For DOHDOT, the value can be GET or POST.
  • statuscode positive integer specifying the expected HTTP status code in the response to an HTTPADV or DOHDOT check.
  • ipv6 boolean specifying the check should run against an ipv6 address - PING, HTTP, HTTPCONTENT, HTTPADV, WHOIS, and DOHDOT checks.
  • regex boolean/string to set whether the 'contentstring' element is a regular expression or just a string to be matched. Can be "true"/true/1 or "false"/false/0. HTTPCONTENT and HTTPADV only.
  • servername string to specify the FQDN sent to SNI services in the SSL check.
  • snmpv string specifying the SNMP version the check should use. Valid values are "1" and "2c". Defaults to "1" - SNMP check only.
  • snmpcom string specifying the SNMP community indicator that should be used. Defaults to 'public' - SNMP check only.
  • verifyvolume boolean to enable the volume detection feature - AUDIO check only.
  • volumemin integer (acceptable range -90 to 0) used by the volume detection feature - AUDIO check only.
  • whoisserver string specifying the WHOIS server FQDN or IPv(4/6) to query - WHOIS check only.

Contact

This is an example declaration of contact resource:

resource "nodeping_contact" "example"{
	custrole = "view"
	name = "Example"
	addresses {
		address = "[email protected]"
		type = "email"
	}
	addresses {
		address = "expl.com"
		type = "webhook"
		action = "PUT"
		data = {"grr": "rrr"}
        headers = {"Content-Type": "application/json"}
        querystrings = {"querykey1":"value1"}
	}
}

The provider closely follows NodePing API, so all parameters like suppressup, suppressfirst and additional parameters for the notification types are available.

  • customer_id a customer id of the subaccount to which the contact belongs. Not needed if the contact is in your primary account.
  • name is the name of the contact, used as a label.
  • custrole set to 'edit,' 'view' or 'notify' to set permissions for this contact. Default is 'view'. Contacts created with an 'edit' or 'view' access level will receive a welcome email message, which by default includes a random password and the suggestion that they change the password immediately. You can avoid these messages by creating the contact with a 'notify' custrole, and then updating them to 'edit' or 'view'. You can also change the content of the welcome email message in the Branding preferences.
  • addresses optional collection of email addresses and phone numbers. See the example above.

Significant difference to the original API is that this terraform provider is aware of the order in with addresses are declared. This is due to the use of address id when connecting checks and addresses. See a more robust example below.

  • priority additional parameter for the pushover notification type. Defaults to 2, with the valid values are 1, 0, -1, or -2 per the priority API documentation for Pushover.

Additional parameters for the webhook notification type:

  • action defaults to 'get' but can be 'put', 'post', 'head', or 'delete'
  • data payload or body of an HTTP POST or PUT request. This can be JSON, XML, or any arbitrary string.
  • headers sets HTTP request headers
  • querystrings sets HTTP query string key/values, that will be appended to your webhook URL as part of the query string.

Contactgroups

This is an example declaration of Contactgroups resource:

resource "nodeping_contact" "example"{
	custrole = "view"
	name = "Example Contact"
	addresses {
		address = "[email protected]"
		type = "email"
	}
}

resource "nodeping_group" "the_group"{
	name = "test"
	members = [nodeping_contact.example.addresses[0].id]
}

The provider follows the Api documentation NodePing API

  • customer_id a customerid of the subaccount the check belongs to.
  • name a name of the contact group, used as a label.
  • members an array of address ids belong to the contact group

Schedule

This is an example declaration of a schedule resource:

resource "nodeping_schedule" "my_schedule"{
	name = "MySchedule"
	data {
		day = "monday"
		time1 = "16:00"
		time2 = "17:00"
		exclude = false
	}
	data {
		day = "sunday"
		allday = true
	}
}

output "my_schedule_name" {
	value = nodeping_schedule.my_schedule.name
}

Note that the data parameter is declared differently than in the official API documentation.

  • customer_id - a customer id of the subaccount to which the schedule belongs.
  • name - this parameter is not present in the official documentation. In practice the API requires an id parameter when creating a new schedule, but then returns a response like this: {"ok":true,"id":"100000000000A0A0A"}, that indicates there is some other id. To avoid this confusion, schedule name is used as its id.
  • data - required object containing the schedule (see the example above). Properties inside each day object are as follows:
    • time1 - start of time span
    • time2 - end of time span
    • exclude - inverts the time span so it is all day except for the time between time1 and time2
    • disabled - disables notifications for this day.
    • allday - enables notifications for the entire day.

More robust example

This example shows how to use all these resources together.

terraform {
  required_providers {
    nodeping = {
      version = "0.0.1"
      source  = "softkraft.co/terraform/nodeping"
    }
  }
}

resource "nodeping_customer" "my_subaccount"{
	name = "NewSubaccount"
	contactname = "Contact Name"
	email =  "[email protected]"
	timezone = "2.0"
	location = "nam"
	emailme = true
	status = "Active"
}

resource "nodeping_contact" "my_contact"{
	customer_id = nodeping_customer.new_subaccount.id
	custrole = "owner"
	name = "MyContact"
	addresses {
		address = "[email protected]"
		type = "email"
	}
	addresses {
		address = "[email protected]"
		type = "email"
	}
}

resource "nodeping_group" "my_group"{
	customer_id = nodeping_customer.my_subaccount.id
	name = "My Group"
	members = [nodeping_contact.my_contact.addresses[0].id, nodeping_contact.my_contact.addresses[1].id]
}

resource "nodeping_schedule" "my_schedule"{
	customer_id = nodeping_customer.my_subaccount.id
	name = "MySchedule"
	data {
		day = "monday"
		time1 = "16:00"
		time2 = "17:00"
		exclude = false
	}
	data {
		day = "sunday"
		allday = true
	}
}

resource "nodeping_check" "my_check"{
	customer_id = nodeping_customer.my_subaccount.id
	label = "MyCheck"
	type = "HTTP"
	target = "http://example.eu/"
	enabled = "inactive"
	notifications {
		contact = nodeping_contact.my_contact.addresses[0].id
		delay = 1
		schedule = "Weekdays"
	}
    notifications {
		contact = nodeping_contact.my_contact.addresses[1].id
		delay = 10
		schedule = nodeping_schedule.my_schedule.name
	}
	homeloc = "false"
	follow = true
	ipv6 = true
}

output "my_customer_id" {
	value = nodeping_customer.my_subaccount.id
}

output "my_check_id" {
	value = nodeping_check.my_check.id
}

output "my_contact_id" {
	value = nodeping_contact.my_contact.id
}

output "my_first_address_id"{
	value = nodeping_contact.my_contact.addresses[0].id
}

output "my_schedule_name" {
	value = nodeping_schedule.my_schedule.name
}

Data sources

Currently three data sources are implemented: nodeping_customer, nodeping_contact, nodeping_group, nodeping_check, nodeping_schedule.

Account

Here's an example use:

data "nodeping_customer" "the_customer" {
	id = var.customer_id
}

output "customer" {
	value = data.nodeping_customer.the_customer
}

Contact

data "nodeping_contact" "my_contact" {
	id = "202103031206A0A0A-A0A0A"
}

output "contact" {
  value = data.nodeping_contact.my_contact
}

Contactgroup

data "nodeping_group" "the_group" {
	id = var.group_id
}

output "group" {
	value = data.nodeping_group.the_group
}

Check

data "nodeping_check" "the_check" {
	id = var.check_id
}

output "check_id" {
	value = data.nodeping_check.the_check.id
}

Schedule

Schedule data source uses name attribute instead of id when running queries.

data "nodeping_schedule" "the_schedule" {
	name = "Weekdays"
}

output "schedule_name" {
	value = data.nodeping_schedule.the_schedule.name
}

output "schedule" {
	value = data.nodeping_schedule.the_schedule
}

More examples

SSL validation

A very simple check for the ssl certificate with notification over an e-mail 10 days before expiry.

esource "nodeping_contact" "first_contact"{
	custrole = "owner"
	name = "Foo Bar"
	addresses {
		address = "[email protected]"
		type = "email"
	}
}

resource "nodeping_check" "ssl_check"{
	label = "SSLExampleCheck"
	type = "SSL"
	target = "https://example.com"
	interval = 1440
	notifications {
		contact = nodeping_contact.first_contact.addresses[0].id
		delay = 1
		schedule = "Days"
	}
	warningdays = 10
}

Development

It will be useful, from the developers' standpoint, to set TF_LOG=DEBUG. More info here.

This project includes a Makefile to ease standard everyday tasks. Currently this includes three commands:

  • make build - builds the package,
  • make install - builds the package, and moves it to terraform plugins,
  • make run_tests - installs the provider, and runs tests.

For make install (and by extension make run_tests) to work, an OS_ARCH environment variable should be set. If it's not present, then "linux_amd64" is assumed.

The make run_tests command requires NODEPING_API_TOKEN environmental variable needs to be set, even if it is declared in terraform files. This is because tests query nodeping API to check that expected resources were created.

Also make run_tests runs tests in parallel, with might make the output unredable. If you want to run tests one by one, simply set -parallel 1 in run_tests code in Makefile.

Terraform keeps a checksum of providers in the project state, so after every plugin re-installation terraform state needs to be reset (removing .terraform, .terraform.lock.hcl, terraform.tfstate).

At this point, the only tests in the project are integration tests. Note that if a test fails with an error, you may be left with some resources created at NodePing. In that case, there might also be a need to restart terraform state.

License

This project is distributed under Apache 2.0 license.

About

Setup AWS Monitoring with Terraform and NodePing

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published