stackit-dev-tools/terraform-provider-stackitprivatepreview
2
0
Fork 0
Code Issues Pull requests 1 Projects Releases Packages Wiki Activity Actions

feat: implement observability log alertgroups (#785)

Browse source
This commit is contained in:
Mauritz Uphoff 2025-04-29 11:43:05 +02:00 • committed by GitHub
parent 616788087c
commit 3c20b7743f
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
15 changed files with 1860 additions and 13 deletions
Download patch file Download diff file Expand all files Collapse all files

4
docs/data-sources/observability_alertgroup.md
View file

@ -3,12 +3,12 @@
page_title: "stackit_observability_alertgroup Data Source - stackit"
subcategory: ""
description: |-
Observability alert group resource schema. Must have a region specified in the provider configuration.
Observability alert group datasource schema. Used to create alerts based on metrics (Thanos). Must have a region specified in the provider configuration.
---
# stackit_observability_alertgroup (Data Source)
Observability alert group resource schema. Must have a `region` specified in the provider configuration.
Observability alert group datasource schema. Used to create alerts based on metrics (Thanos). Must have a `region` specified in the provider configuration.
## Example Usage

47
docs/data-sources/observability_logalertgroup.md Normal file
View file

@ -0,0 +1,47 @@
---
# generated by https://github.com/hashicorp/terraform-plugin-docs
page_title: "stackit_observability_logalertgroup Data Source - stackit"
subcategory: ""
description: |-
Observability log alert group datasource schema. Used to create alerts based on logs (Loki). Must have a region specified in the provider configuration.
---
# stackit_observability_logalertgroup (Data Source)
Observability log alert group datasource schema. Used to create alerts based on logs (Loki). Must have a `region` specified in the provider configuration.
## Example Usage
```terraform
data "stackit_observability_logalertgroup" "example" {
project_id = "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
instance_id = "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
name = "example-log-alert-group"
}
```
<!-- schema generated by tfplugindocs -->
## Schema
### Required
- `instance_id` (String) Observability instance ID to which the log alert group is associated.
- `name` (String) The name of the log alert group. Is the identifier and must be unique in the group.
- `project_id` (String) STACKIT project ID to which the log alert group is associated.
### Read-Only
- `id` (String) Terraform's internal resource ID. It is structured as "`project_id`,`instance_id`,`name`".
- `interval` (String) Specifies the frequency at which rules within the group are evaluated. The interval must be at least 60 seconds and defaults to 60 seconds if not set. Supported formats include hours, minutes, and seconds, either singly or in combination. Examples of valid formats are: '5h30m40s', '5h', '5h30m', '60m', and '60s'.
- `rules` (Attributes List) (see [below for nested schema](#nestedatt--rules))
<a id="nestedatt--rules"></a>
### Nested Schema for `rules`
Read-Only:
- `alert` (String) The name of the alert rule. Is the identifier and must be unique in the group.
- `annotations` (Map of String) A map of key:value. Annotations to add or overwrite for each alert
- `expression` (String) The LogQL expression to evaluate. Every evaluation cycle this is evaluated at the current time, and all resultant time series become pending/firing alerts.
- `for` (String) Alerts are considered firing once they have been returned for this long. Alerts which have not yet fired for long enough are considered pending. Default is 0s
- `labels` (Map of String) A map of key:value. Labels to add or overwrite for each alert

199
docs/guides/ske_log_alerts.md Normal file
View file

@ -0,0 +1,199 @@
---
page_title: "SKE Log Alerts with STACKIT Observability"
---
# SKE Log Alerts with STACKIT Observability
## Overview
This guide walks you through setting up log-based alerting in STACKIT Observability using Promtail to ship Kubernetes logs.
1. **Set Up Providers**
Begin by configuring the STACKIT and Kubernetes providers to connect to the STACKIT services.
```hcl
provider "stackit" {
region = "eu01"
}
provider "kubernetes" {
host = yamldecode(stackit_ske_kubeconfig.example.kube_config).clusters.0.cluster.server
client_certificate = base64decode(yamldecode(stackit_ske_kubeconfig.example.kube_config).users.0.user.client-certificate-data)
client_key = base64decode(yamldecode(stackit_ske_kubeconfig.example.kube_config).users.0.user.client-key-data)
cluster_ca_certificate = base64decode(yamldecode(stackit_ske_kubeconfig.example.kube_config).clusters.0.cluster.certificate-authority-data)
}
provider "helm" {
kubernetes {
host = yamldecode(stackit_ske_kubeconfig.example.kube_config).clusters.0.cluster.server
client_certificate = base64decode(yamldecode(stackit_ske_kubeconfig.example.kube_config).users.0.user.client-certificate-data)
client_key = base64decode(yamldecode(stackit_ske_kubeconfig.example.kube_config).users.0.user.client-key-data)
cluster_ca_certificate = base64decode(yamldecode(stackit_ske_kubeconfig.example.kube_config).clusters.0.cluster.certificate-authority-data)
}
}
```
2. **Create SKE Cluster and Kubeconfig Resource**
Set up a STACKIT SKE Cluster and generate the associated kubeconfig resource.
```hcl
resource "stackit_ske_cluster" "example" {
project_id = "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
name = "example"
kubernetes_version = "1.31"
node_pools = [
{
name = "standard"
machine_type = "c1.4"
minimum = "3"
maximum = "9"
max_surge = "3"
availability_zones = ["eu01-1", "eu01-2", "eu01-3"]
os_version_min = "4081.2.1"
os_name = "flatcar"
volume_size = 32
volume_type = "storage_premium_perf6"
}
]
maintenance = {
enable_kubernetes_version_updates = true
enable_machine_image_version_updates = true
start = "01:00:00Z"
end = "02:00:00Z"
}
}
resource "stackit_ske_kubeconfig" "example" {
project_id = "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
cluster_name = stackit_ske_cluster.example.name
refresh = true
}
```
3. **Create Observability Instance and Credentials**
Establish a STACKIT Observability instance and its credentials to handle alerts.
```hcl
locals {
alert_config = {
route = {
receiver = "EmailStackit",
repeat_interval = "1m",
continue = true
}
receivers = [
{
name = "EmailStackit",
email_configs = [
{
to = "<email>"
}
]
}
]
}
}
resource "stackit_observability_instance" "example" {
project_id = "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
name = "example"
plan_name = "Observability-Large-EU01"
alert_config = local.alert_config
}
resource "stackit_observability_credential" "example" {
project_id = "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
instance_id = stackit_observability_instance.example.instance_id
}
```
4. **Install Promtail**
Deploy Promtail via Helm to collect logs and forward them to the STACKIT Observability Loki endpoint.
```hcl
resource "helm_release" "promtail" {
name = "promtail"
repository = "https://grafana.github.io/helm-charts"
chart = "promtail"
namespace = kubernetes_namespace.monitoring.metadata.0.name
version = "6.16.4"
values = [
<<-EOF
config:
clients:
# To find the Loki push URL, navigate to the observability instance in the portal and select the API tab.
- url: "https://${stackit_observability_credential.example.username}:${stackit_observability_credential.example.password}@<your-loki-push-url>/instances/${stackit_observability_instance.example.instance_id}/loki/api/v1/push"
EOF
]
}
```
5. **Create Alert Group**
Create a log alert that triggers when a specific pod logs an error message.
```hcl
resource "stackit_observability_logalertgroup" "example" {
project_id = "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
instance_id = stackit_observability_instance.example.instance_id
name = "TestLogAlertGroup"
interval = "1m"
rules = [
{
alert = "SimplePodLogAlertCheck"
expression = "sum(rate({namespace=\"example\", pod=\"logger\"} |= \"Simulated error message\" [1m])) > 0"
for = "60s"
labels = {
severity = "critical"
},
annotations = {
summary : "Test Log Alert is working"
description : "Test Log Alert"
},
},
]
}
```
6. **Deploy Test Pod**
Launch a pod that emits simulated error logs. This should trigger the alert if everything is set up correctly.
```hcl
resource "kubernetes_namespace" "example" {
metadata {
name = "example"
}
}
resource "kubernetes_pod" "logger" {
metadata {
name = "logger"
namespace = kubernetes_namespace.example.metadata[0].name
labels = {
app = "logger"
}
}
spec {
container {
name = "logger"
image = "bash"
command = [
"bash",
"-c",
<<EOF
while true; do
sleep $(shuf -i 1-3 -n 1)
echo "ERROR: $(date) - Simulated error message $(shuf -i 1-100 -n 1)" 1>&2
done
EOF
]
}
}
}
```

4
docs/resources/observability_alertgroup.md
View file

@ -3,12 +3,12 @@
page_title: "stackit_observability_alertgroup Resource - stackit"
subcategory: ""
description: |-
Observability alert group resource schema. Must have a region specified in the provider configuration.
Observability alert group resource schema. Used to create alerts based on metrics (Thanos). Must have a region specified in the provider configuration.
---
# stackit_observability_alertgroup (Resource)
Observability alert group resource schema. Must have a `region` specified in the provider configuration.
Observability alert group resource schema. Used to create alerts based on metrics (Thanos). Must have a `region` specified in the provider configuration.
## Example Usage

80
docs/resources/observability_logalertgroup.md Normal file
View file

@ -0,0 +1,80 @@
---
# generated by https://github.com/hashicorp/terraform-plugin-docs
page_title: "stackit_observability_logalertgroup Resource - stackit"
subcategory: ""
description: |-
Observability log alert group resource schema. Used to create alerts based on logs (Loki). Must have a region specified in the provider configuration.
---
# stackit_observability_logalertgroup (Resource)
Observability log alert group resource schema. Used to create alerts based on logs (Loki). Must have a `region` specified in the provider configuration.
## Example Usage
```terraform
resource "stackit_observability_logalertgroup" "example" {
project_id = "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
instance_id = "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
name = "example-log-alert-group"
interval = "60m"
rules = [
{
alert = "example-log-alert-name"
expression = "sum(rate({namespace=\"example\", pod=\"logger\"} |= \"Simulated error message\" [1m])) > 0"
for = "60s"
labels = {
severity = "critical"
},
annotations = {
summary : "example summary"
description : "example description"
}
},
{
alert = "example-log-alert-name-2"
expression = "sum(rate({namespace=\"example\", pod=\"logger\"} |= \"Another error message\" [1m])) > 0"
for = "60s"
labels = {
severity = "critical"
},
annotations = {
summary : "example summary"
description : "example description"
}
},
]
}
```
<!-- schema generated by tfplugindocs -->
## Schema
### Required
- `instance_id` (String) Observability instance ID to which the log alert group is associated.
- `name` (String) The name of the log alert group. Is the identifier and must be unique in the group.
- `project_id` (String) STACKIT project ID to which the log alert group is associated.
- `rules` (Attributes List) Rules for the log alert group (see [below for nested schema](#nestedatt--rules))
### Optional
- `interval` (String) Specifies the frequency at which rules within the group are evaluated. The interval must be at least 60 seconds and defaults to 60 seconds if not set. Supported formats include hours, minutes, and seconds, either singly or in combination. Examples of valid formats are: '5h30m40s', '5h', '5h30m', '60m', and '60s'.
### Read-Only
- `id` (String) Terraform's internal resource ID. It is structured as "`project_id`,`instance_id`,`name`".
<a id="nestedatt--rules"></a>
### Nested Schema for `rules`
Required:
- `alert` (String) The name of the alert rule. Is the identifier and must be unique in the group.
- `expression` (String) The LogQL expression to evaluate. Every evaluation cycle this is evaluated at the current time, and all resultant time series become pending/firing alerts.
Optional:
- `annotations` (Map of String) A map of key:value. Annotations to add or overwrite for each alert
- `for` (String) Alerts are considered firing once they have been returned for this long. Alerts which have not yet fired for long enough are considered pending. Default is 0s
- `labels` (Map of String) A map of key:value. Labels to add or overwrite for each alert