Due to billing or organizational structures, certain parts of the infrastructure could be divided over several Azure subscriptions. From an infrastructure management point of view however, it might be interesting to manage the resources in those multiple subscriptions in one Terraform playbook. In the required_providers section, the configuration_aliases must be configured first (usually in the…
Ansible provides a more flexible way to fine-tune Hashicorp Packer images compared to cloud-init. Playbooks can be executed once the guest image building is ready and boots up for the first time. This allows to create different types of Packer images based on different playbooks. In this article, Packer images will created for Azure using…
De-duplication of configuration information is key when managing large environments which use different types of automation (Terraform, Jenkins, Ansible, scripts executed as Systemd timers, Puppet…). Although many different configuration management tools exist (RDBMS, Consul, …), one of the easiest to use is Hiera or just a plain normal Git repository with YAML files, in some…
Libvirt (libvirtd) nodes (based on KVM and Qemu) are a great and cheap (read: free) alternative of deploying virtual nodes in a cloud. Required is a server which will act as a hypervisor, in our article we chose to use a Hetzner server installed with Ubuntu Linux 20.4-lts. After the default installation of Ubuntu 20.4-lts,…
The subnets accessor in the azurerm_virtual_network Terraform data source returns a list of subnet names only. In most cases however, you will need to use a or multiple subnet IDs, for instance when deploying virtual machines. Instead of creating a new datasource (for possibly a small list of subnets) for each virtual machine you want…
