gsctl v0.18.1

New features

• The gsctl show cluster command now also prints the URL of this cluster's details in our web UI. In terminals like iTerm (Mac) this allows you to open the web UI page with one Cmd-Click. (#481)

Bugfixes

• Fixed some incorrect usage info in the gsctl update cluster command. (#482)

gsctl v0.18.0

This is the first gsctl release fully supporting node pool clusters.

Improvements

Most improvements have been included in the 0.18.x pre-releases (here and there). Please check the release notes for details.

Additionally, this release brings:

• gsctl create nodepool: fix handling of minimum and maximum worker node count if only one of the two are given. (#479)
• Handling of a Malformed response error, which is normally a permission problem due to bad network membership (e. g. missing VPN connection), has been improved in all commands. (#478)
• gsctl show clusters: For a node pools cluster, we added some guidance to the output to help you discover node pools details etc. (#480)

gsctl v0.17.5

Improvements

• When deleting a cluster using the gsctl delete cluster command, you will be prompted to enter the ID of the cluster to be deleted. The goal here is to help you re-consider which cluster you are about to destroy forever. (#463)
• The gsctl list clusters command can now show when there are cluster in deletion. However this only applies to v5 clusters, not v4, and will thus only affect AWS at first. (#468, #471)
• The gsctl create cluster command now gives more specific error details when a cluster definition read from STDIN cannot be parsed. (#465)
• The handling of an API error with status code 500 (internal server error) has been aligned and made more user-friendly in all commands. (#461)

Bugfixes

• A problem with the gsctl scale cluster command has been fixed that would make the command effectively unusable. (#474)
• Usage info in gsctl create cluster has been fixed. (#469, #464)

gsctl v0.18.0 RC3

New features

• The gsctl upgrade cluster command now supports v5 clusters. In case you wonder, those are the ones supporting node pools. (#470)

Improvements

• When deleting a cluster using the gsctl delete cluster command, you will be prompted to enter the ID of the cluster to be deleted. The goal here is to help you re-consider which cluster you are about to destroy forever. (#463)
• The gsctl list clusters command can now show when there are cluster in deletion. However this only applies to v5 clusters, not v4, and will thus only affect AWS at first. (#468, #471)
• The gsctl list nodepools command now shows a friendly error when the given cluster is not a node pools (v5) cluster. (#466)
• The gsctl create cluster command now gives more specific error details when a cluster definition read from STDIN cannot be parsed. (#465)
• The handling of an API error with status code 500 (internal server error) has been aligned and made more user-friendly in all commands. (#461)
• The error handling for gsctl delete nodepool has been improved. (#467)

Bugfixes

• A problem with the gsctl scale cluster command has been fixed that would make the command effectively unusable. (#474)
• Usage info in gsctl create cluster has been fixed. (#469, #464)

gsctl v0.18.0 RC1

This is a pre-release for 0.18.0 which introduces Node pools functionality to gsctl. For more information on node pools, see https://docs.giantswarm.io/basics/nodepools/

New features

• The gsctl create cluster command supports clusters with node pools. Note: Many flags which allowed to specify worker node details in the past have been removed. To specify worker node details, please use the cluster definition format from now on.
• gsctl list nodepools allows listing node pools of a cluster.
• gsctl create nodepool allows creating a node pool.
• gsctl update nodepool allows scaling and renaming of a node pool.
• gsctl delete nodepool does what the name indicates.

We welcome your feedback! If you find anything unexpected, please let us know in your Slack channel or create an issue.

gsctl v0.17.4

New features

• The gsctl update cluster command has been added to allow renaming of a cluster. (#455)

gsctl v0.17.3

New features

• The gsctl list releases command now allows to print output as JSON, to simplify the use in automation and scripting. Use the --output json (short: -o json) flag to enable JSON output. #451
• The gsctl info command now prints environment variables relevant to gsctl. #450

Bugfixes

• Enabled the use of gsctl show cluster again on Azure and KVM installations. #454
• Fixed a bug that would pretty much make the output of gsctl list releases unusable. #453

gsctl v0.17.2

Bugfixes

• gsctl create kubeconfig: add missing error handling when switching between V4 and V5 api endpoints. (#449)

gsctl v0.17.1

New features

• The commands gsctl list clusters and gsctl list keypairs now allow to print output as JSON, to simplify the use in automation and scripting. Use the --output json (short: -o json) flag to enable JSON output. (#431, #432)

Minor changes

• The timeout for most API calls has been set to 10 seconds. Before it was 60 seconds. This should reduce your wait time until you find out something is wrong with your connection. (#445)

Bugfixes

• If no releases can be found, gsctl list releases no longer prints table headers for an empty table. (#446)

gsctl v0.17.0

Minor changes

• Deprecated flags have been removed from gsctl scale cluster: --memory-gb, --num-cpus,
  --storage-gb. (#442)
• The gsctl create kubeconfig command, when used with the --self-contained flag, would log two lines of information which were not useful. This has been disabled as part of (#437).

Bugfixes

• A problem in gsctl scale cluster has been fixed that caused a panic with error Provider must not be empty: invalid config error. (#442)