Dynatrace Availability Dashboards

Purpose

The purpose of this repository is to automate the creation of Synthetic Monitors, Management Zones, SLOS and Dashboards within Dynatrace, by querying Ingress objects in each environment on the CFT Kubernetes clusters

What's inside?

• Terraform
  Components defined in Terraform to build associated infrastructure
• Python and Bash scripts
  These scripts are used to:
  • Autogenerate yaml files per-environment based on Ingress objects withiin clusters, to update Terraform configuration files
  • Create Pull Requests if there are changes to Ingress objects within clusters
• Azure pipelines
  • Daily - Runs each morning at 7AM to fetch latest changes
  • Deployment - Deploy infrastructure changes and update dahsboards
• Github Actions
  We use Github Actions to run:
  • Pre-Commit checks
  • Unit Tests
• Renovate
  Renovate is configured to keep our dependencies up-to-date and keep a running dashboard

Terraform docs

Requirements

Name	Version
terraform	1.4.0
dynatrace	1.21.0

Providers

Name	Version
dynatrace	1.21.0
time	n/a

Resources

Name	Type
dynatrace_dashboard.availability	resource
dynatrace_http_monitor.availability	resource
dynatrace_maintenance.auto_shutdown_blackout	resource
dynatrace_maintenance.auto_shutdown_weekends	resource
dynatrace_maintenance.planned_blackout	resource
dynatrace_management_zone.availability	resource
dynatrace_slo.availability	resource
time_sleep.management_zones	resource

Inputs

Name	Description	Type	Default	Required
auto_window_end_date	This is the date that the cluster-shutdown maintenance window (stopped alerting) should end	string	"2030-01-01"	no
auto_window_end_time	This is the time that the cluster-shutdown maintenance window (stopped alerting) should end	string	"08:30:00"	no
auto_window_start_date	This is the date that the cluster-shutdown maintenance window (stopped alerting) should begin	string	"2023-01-31"	no
auto_window_start_time	This is the time that the cluster-shutdown maintenance window (stopped alerting) should begin	string	"20:00:00"	no
automated_weekday_maintenance	This is a toggle to enable or disable the automatic maintenance window for an environment on weekdays, defaults to true as we have cluster shutdowns.	bool	true	no
automated_weekend_maintenance	This is a toggle to enable or disable the automatic maintenance window for an environment on the weekend, defaults to true as we have cluster shutdowns.	bool	true	no
builtFrom	builtFrom variable required by cnp terraform template. Not in use by terraform	string	null	no
dt_env_url	Dynatrace environment URL	string	n/a	yes
dynatrace_platops_api_key	Dynatrace API access key	string	n/a	yes
env	Name of the environment to build in	string	n/a	yes
planned_maintenance	This is a toggle to enable or disable a planned maintenance window for an environment	bool	false	no
planned_window_end_time	End time of the planned maintenance window	string	"2023-01-30T20:00:00"	no
planned_window_start_time	Start time of the planned maintenance window	string	"2023-01-30T19:00:00"	no
product	product variable required by cnp terraform template. Not in use by terraform	string	null	no
weekend_window_end_time	This is the time that the cluster-shutdown maintenance window (stopped alerting) should end during the weekend	string	"23:59:59"	no
weekend_window_start_time	This is the time that the cluster-shutdown maintenance window (stopped alerting) should begin during the weekend	string	"00:00:00"	no

Outputs

Name	Description
ado_dashboard_url	Used to display created dashboard URLs as vso[task.logissue type=warning]
dashboard_url	Dynatrace Dashboard URL

Blackouts (Maintenance Windows)

There are two types of Maintenance Windows being handled by this repository:

• Automated
  • Created here and used to turn off our HTTP monitors and alerting for the nonprod environments that are shut down during the evenings, and restarted in the morning. To change the settings for a specific environment, edit the {env}.tfvars file.
• Manual
  • Created here and should be enabled in a case where we have a planned suspected outage for one or more environments, i.e a cluster switchover. To achieve this:
    • Set the enabled flag to true in a given {env}.tfvars file, for example in demo
    • Update the variables defined a given {env}.tfvars file file. This allows you to define the start, and end time of the planned outage within that environment.