v3.0.0

Changes

 

• implemented breaking feature: Terraform Azure Remote State Backend
• updated terraform to 0.12.9
• updated runtime to ksandermann/cloud-toolbox:2019-09-19_01

Bootstrap environment

 
ksandermann/cloud-toolbox:2019-09-19_01
Terraform: 0.12.9
recommended Docker: 19.03.2

v2.0.0

Changes

 

• changed node_resource_group for aks cluster from cluster_fqdn to platform_rg_name + "_ClusterResources"
• marked secrets and AKS kubeconfig outputs as sensitive
• updated runtime to ksandermann/cloud-toolbox:2019-09-17_01
• bumped terraform to 0.12.8
• bumped azure provider versions to latest supported versions
• added terraform output for kube configs to apply script
• bumped example cluster AKS version to 1.13.10
• implemented output var with the Azure DNS zone nameservers
• implemented probe_id for lb rules, fixing #1
• modularized destroy and plan scripts
• added set -e to all scripts again
• implemented mulitple output vars for azure provider - private IPs for cluster vms, public IPs for bastions
• added tenant_id to aks ad integration
    

Bootstrap environment

 
ksandermann/cloud-toolbox:2019-09-17_01
Terraform: 0.12.8
recommended Docker: 19.03.2

v1.1.1

Changes

 

• AKS Provider: merged dns and cluster modules
• AKS Provider: migrated loadbalancer public IP to clusterresources rg
    

Bootstrap environment

 
ksandermann/cloud-toolbox:2019-08-08_01
Terraform: 0.12.6
recommended Docker: 18.09.2

v1.1.0

Changes

 

• implemented support for AKS to fully bootstrap an AKS Kubernetes Cluster
    

Bootstrap environment

 
ksandermann/cloud-toolbox:2019-08-08_01
Terraform: 0.12.6
recommended Docker: 18.09.2

v1.0.0

Changes

• enabled ip-forwarding for all nics
• added empty azure route for cluster subnet
• added message that DNS zone will not get removed when running destroy
• added documentation for kubespray pitfalls
• adjusted kubespray sample inventories

Bootstrap environment

ksandermann/cloud-toolbox:2019-07-18_01
Terraform: 0.12.4
recommended Docker: 18.09.2

v0.2.1

v0.2.1

Changes

• changed default vm sizes to support Premium Storage
• destroy now keeping the dns zone

Bootstrap environment

• ksandermann/cloud-toolbox:2019-07-18_01
• Terraform: 0.12.4
• recommended Docker: 18.09.2

v0.2.0

v0.2.0

Changes

• moved from multi-zone high-availability to availability sets (multiple racks in one region).
  As it turns out, Azure does not support one of the following patterns:

  1. Adding blob-based disks from a storage account to a VM that resists in a specific zones.
  2. Adding both blob-based disks and managed disk to the same VM at the same time.
  3. Creating disks that reside in multiple zones at the same time (zone-redundancy)
  4. Automatic migration of a disk from one zone to another

  This causes the problem, that the Kubernetes Azure Cloud Provider is not able to properly create/manage disks:

1. The cloud provider has Azure availability-zones support since 1-12 as stated here. Though, this feature is not reasonable to use, as this only uses node affinities/constraints to add disks only to nodes
   that reside in the same zone as the disk. In case of a zone failure, the disk is also lost. Therefore, the whole point of using multiple zones in order to be safe against zone failure, is not met.
2. OpenShift 3.11 still uses Kubernetes 1.11 and does not have the above feature at all.

Bootstrap environment

• ksandermann/cloud-toolbox:2019-07-18_01
• Terraform: 0.12.4
• recommended Docker: 18.09.2

v0.1.1

v0.1.1

hotfixes:

• fixed documentation from deploy.sh to apply.sh

• fixed bug: A-record from bastion now pointing to public instead of private IP

Bootstrap environment:

• ksandermann/cloud-toolbox:2019-07-18_01
• Terraform: 0.12.4
• recommended Docker: 18.09.2

v0.1.0

v0.1.0 - finally opensourced this

• Bootstrap environment: ksandermann/cloud-toolbox:2019-07-18_01
• Terraform: 0.12.4
• recommended Docker: 18.09.2