A Terraform plugin that provides resources for Datadog.
It currently supports 4 resources based on the Datadog monitor originally contributed by Vincenzo Prignano of Segmentio.
-
Datadog Monitors: datadog_monitor, this monitor supports service checks, metric, outlier and event alerts.
-
Service Checks: datadog_service_check. Does not detect upstream changes, and will be removed.
-
Metric Alerts: datadog_metric_alert. Does not detect upstream changes, and will be removed.
-
Outlier Alerts: datadog_outlier_alert, see introducing-outlier-detection-in-datadog. Does not detect upstream changes, and will be removed.
Feel free to open new issues for extra resources or bugs you find.
Want to contribute? Find a resource you want to add or work on an issue over here.
Download builds for Darwin, Linux and Windows from the releases page. Pre-release is rebuild on each merge to master.
This plugin will create a monitor. By default it will monitor reports from all hosts or other sources that run a given check or report a certain metric.
The flow would be:
- Run service checks, or report metrics from host(s)
- Create monitors for these using this plugin
Filter which hosts / groups are monitored by using tags.
Example configuration:
resource "datadog_monitor" "foo" {
name = "name for monitor foo"
type = "metric alert" // Alternatives: "service check", "event alert" or "query alert" (undocumented, outlier alert)
message = "A message Notify: @hipchat-channel"
escalation_message = "An escalation message @pagerduty"
query = "avg(last_1h):avg:aws.ec2.cpu{environment:foo,host:bar} by {host} > 3"
thresholds {
ok = 0
warning = 1
critical = 3
}
notify_no_data = true
renotify_interval = 40
notify_audit = true
timeout_h = 70
include_tags = false
silenced {
"*" = 0
}
Deprecated, this check does not update state, use the generic monitor alert.
This plugin will create a monitor, but not a service check. By default it will monitor reports from all hosts that run a given service check.
Example configuration:
resource "datadog_service_check" "foo" {
name = "name for service check foo"
message = <<EOF
{{#is_alert}}Service check foo is critical{/is_alert}}
{{#is_warning}}Service check foo is at warning level{{/is_warning}}
{{#is_recovery}}Service check foo has recovered{{/is_recovery}}
Notify: @hipchat-channel
EOF
check = "datadog.agent.up"
check_count = 3
tags = ["environment:foo", "host:bar"]
keys = ["foo", "bar"]
thresholds {
ok = 1 // Optional
warning = 2 // Optional
critical = 3 // Required, formally known as "threshold"
}
notify_no_data = false
}Deprecated, this check does not update state, use the generic monitor alert.
Example configuration:
resource "datadog_metric_alert" "bar" {
name = "TF: bar"
message = <<EOF
{{#is_alert}}Metric alert check bar is critical{/is_alert}}
{{#is_warning}}Metric alert bar is warning{{/is_warning}}
{{#is_recovery}}Metric alert bar has recovered{{/is_recovery}}
Notify: @hipchat-channel
EOF
metric = "datadog.dogstatsd.packet.count"
tags = ["*"]
keys = ["host"]
time_aggr = "avg"
time_window = "last_1h"
space_aggr = "sum"
// the `query` parameter can also be used instead of the above parameters
// `query = “avg(last_1h):sum:datadog.dogstatsd.packet.count{*} by {host}”`
// is equivalent to specifying the above parameters
// if `query` is used `metric`, `tags`, `keys`, `time_aggr`, `time_window` and `space_aggr` must not be used
operator = ">"
notify_no_data = 0
thresholds {
ok = 1 // Optional
warning = 2 // Optional
critical = 3 // Required, formally known as "threshold"
}
}Deprecated, this check does not update state, use the generic monitor alert.
Example configuration:
resource "datadog_outlier_alert" "foo" {
name = "name for outlier_alert foo"
message = "description for outlier_alert @hipchat-channel"
algorithm = "mad"
metric = "system.load.5"
tags = ["environment:foo", "host:foo"]
keys = ["host"]
time_aggr = "avg" // avg, sum, max, min, change, or pct_change
time_window = "last_1h" // last_#m (5, 10, 15, 30), last_#h (1, 2, 4), or last_1d
space_aggr = "avg" // avg, sum, min, or max
threshold = 3.0 // tolerance
notify_no_data = false
}Like any other Terraform interactions.
Tip: export DATADOG_API_KEY and DATADOG_APP_KEY as environment variables.
###Plan
ojongerius@hipster ~/dev/go/datadog terraform plan
Refreshing Terraform state prior to plan...
The Terraform execution plan has been generated and is shown below.
Resources are shown in alphabetical order for quick scanning. Green resources
will be created (or destroyed and then created if an existing resource
exists), yellow resources are being changed in-place, and red resources
will be destroyed.
Note: You didn't specify an "-out" parameter to save this plan, so when
"apply" is called, Terraform can't guarantee this is what will execute.
+ datadog_metric_alert.statsd_packet_count
keys.#: "0" => "1"
keys.0: "" => "host"
message: "" => "statsd packet count {{host.hostname}}"
metric: "" => "datadog.dogstatsd.packet.count"
name: "" => "TF: stats_packet_count"
notify: "" => "@ojongerius@warning.com"
notify_no_data: "" => "0"
operator: "" => ">"
renotify_interval: "" => "0"
space_aggr: "" => "sum"
tags.#: "0" => "1"
tags.0: "" => "*"
threshold: "" => "2"
time_aggr: "" => "avg"
time_window: "" => "last_1h"
Plan: 1 to add, 0 to change, 0 to destroy.###Apply
ojongerius@hipster ~/dev/go/datadog terraform apply
datadog_metric_alert.statsd_packet_count: Creating...
keys.#: "0" => "1"
keys.0: "" => "host"
message: "" => "statsd packet count {{host.hostname}}"
metric: "" => "datadog.dogstatsd.packet.count"
name: "" => "TF: stats_packet_count"
notify: "" => "@ojongerius@warning.com"
notify_no_data: "" => "0"
operator: "" => ">"
renotify_interval: "" => "0"
space_aggr: "" => "sum"
tags.#: "0" => "1"
tags.0: "" => "*"
threshold: "" => "2"
time_aggr: "" => "avg"
time_window: "" => "last_1h"
datadog_metric_alert.statsd_packet_count: Creation complete
Apply complete! Resources: 1 added, 0 changed, 0 destroyed.
The state of your infrastructure has been saved to the path
below. This state is required to modify and destroy your
infrastructure, so keep it safe. To inspect the complete state
use the `terraform show` command.
State path: terraform.tfstate> make test
go generate ./...
TF_ACC= go test ./... -timeout=30s -parallel=4
? github.com/ojongerius/terraform-provider-datadog[no test files]
ok github.com/ojongerius/terraform-provider-datadog/datadog0.007s
go tool vet -asmdecl -atomic -bool -buildtags -copylocks -methods -nilfunc
-printf -rangeloops -shift -structtags -unsafeptr .Much more extensive but need a valid Datadog API and APP key.
These tests will both create and destroy real resources.
ojongerius@hipster ~/gocode/src/github.com/ojongerius/terraform-provider-datadog KISS ● make
testacc
go generate ./...
TF_ACC=1 go test ./datadog -v -timeout 90m
=== RUN TestProvider
--- PASS: TestProvider (0.00s)
=== RUN TestProvider_impl
--- PASS: TestProvider_impl (0.00s)
=== RUN TestAccDatadogMetricAlert_Basic
--- PASS: TestAccDatadogMetricAlert_Basic (3.65s)
=== RUN TestAccDatadogOutlierAlert_Basic
--- PASS: TestAccDatadogOutlierAlert_Basic (2.46s)
=== RUN TestAccDatadogServiceCheck_Basic
--- PASS: TestAccDatadogServiceCheck_Basic (2.49s)
PASS
ok github.com/ojongerius/terraform-provider-datadog/datadog 8.614sBuilding defaults to the platform you run on, and depends on gox. If you do not have it installed:
go get github.com/mitchellh/gox> make bin
go generate ./...
Compiling for OS: darwin and ARCH: amd64
Number of parallel builds: 8
--> darwin/amd64: github.com/ojongerius/terraform-provider-datadog
Looking for Terraform install
Moving terraform-provider-datadog_darwin_amd64 to
/Applications/terraform/terraform-provider-datadog
Resulting binary:
-rwxr-xr-x 1 ojongerius staff 10442740 4 Aug 18:32
/Applications/terraform/terraform-provider-datadog