Releases: openshift/rosa
Releases · openshift/rosa
Release v1.2.14
- minor refactor to improve code clarity wrt addons
- fix: reprompt user when passed invalid addon argument
- fix: permissions boundary shouldn't be asked if supposed to skip interactive
- feat: add force param to forcefully ensure policies
- Edit machinepool min replicas
- fix: incorrect flags in message and hidden for upgrade roles
- IDP related minor cleanup
- Add port for OAuth Callback URI in OpenID
- fix: only show root failure if it is not a suggestion error
- Refactor
create account roles command
to use interfaces - fix: add region when creating manual s3 bucket for oidc config
- feat: add user prefix to oidc configuration
- feat: add spinner creating oidc config
- fix: show info report when deleting operator roles
- fix: oidc endpoint url should be of https scheme
- cmd/dlt/machinepool: add confirm flag
- feat: Add command delete oidc-config and minor fixes
- fix: add question for private key secret arn
Release v1.2.13
- Skip region check if we use shard pinning
- Remove GitHub IDP dependency to console availability
- Bump ocm-sdk-go to v0.1.310
- Fix managed policies cluster upgrade path
- Delete roles with managed policies
- feat: store private key for byo oidc in secrets manager
- [SDA-7757] byo OIDC secret arn support (#1018)
- feat: add download rosa option
- Create OCM role with managed policies
- Upgrade to Go 1.19
- Delete OCM role with managed policies
- feat: add force param to forcefully ensure policies
Release v1.2.12
- fix: Incorrect OIDC Provider Sometimes Targeted for Deletion
- Removed len(tags) > 0 from if for Tags question in interactive
- Revert "[SDA-7662] Display Tags question in interactive mode"
- fix: check current values to see if there are no real changes
- Re-added tags question
- Add the default-mp-labels flag to cluster create command on interactive mode
- Fix OAuth callback URL logic for HyperShift clusters
- Now updating stsBuilder.AutoMode properly
- Added constant for "auto" string
- Upgrade operator roles with managed policies
- feat: command create oidc-config
- feat: add manual mode
- Edited query for GetClusterUsingSubscription to fix deletions
- Fix tags passing in cluster creation interactive mode
- fix: using ARNValidator instead of arn Parse when parsed is not used
- Upgrade roles command - handle managed policies
- fix: only checking '"' character and leaving regex validation for CS
- fix: delete correct operator roles and support hypershift
- fix: accept pre release version during upgrade
- Ensure console URL is available before offering it to the user
- feat: consider current version incompatible
- SDCICD-893: cmd/create/machinepool: support output flag (#1014)
Release v1.2.11
- fix: check if any new operator roles have been created
- fix: spacing listing users
- fix: remove auth url info from LDAP idp when listing
- Upgrade to ocm-sdk-go 0.1.299
- fix: upgrading from pre release would fail to validate version
- Log a warning if the user's organization doesn't have the needed capability
- fix: using unified path on upgrade roles/operator-roles
- feat: validates machine pool label
- fix: add error message when CA is passed but github hostname is not
- Describe cluster - print
infra_id
to the output - fix: using lower case before comparing expected acc role arns
- feat: retrieve operator role prefix from backend
- fix: Accomodate inline policies in new upgrade roles flow
- feat: using LCP to retrieve operator policy prefix
- Adjust NodePool headers
- Allow editing default machine pool labels
- Enable day1 default machine pool labels
- fix: hide region arg in account roles commands
- fix: removing local contains in favor of helper.Contains
- fix: Use default/in-place value for addon param first
- Add default machine pool labels validations
- make rosa describe upgrade
- fix: remove channel group from recreate output, this is treated within creation flow
- Bump OCM SDK GO version to v0.1.303
- Refactor
GetPolicies
function - fix ux issues related to rosa describe
- fix: using tabwritters options instead of manually formatting
- Fix scaling bug and improve interactive mode
- Create account roles with existing policies
- Fix bug - create managed account roles
- Refactor get policy details and ARN
- Create operator role with existing policies
- Fix linter errors - add constant for string "true"
- chore: refactor sort strings helper
- feat: hide region from other globally available commands
Release v1.2.10
- Check HostedCP version support also in interactive mode + align versions
- fix: going through all policies to check upgrade
Release v1.2.9
This is a development release. Use 1.2.10 instead for live environments.
- feat: set interactive enabled if local flags are unchanged, except for cluster flag
- feat: don't send update request if there are no changes
- refactor: setting up a function to look into given params expected to be unchanged
- Hosted-cp flag now forces byo vpc prompt
- bump ocm sdk to 0.1.289
- update hcp naming convention for latest SDK
- add market specific billing options for addon installations
- [SDA-6984] Add support for nightly builds for hcp
- Filter regions where hcp is avalaible
- feat: check if rosa cli is up to date
- fix: lint
- refactor: clean up
- fix: favor replicas instead of deprecated compute-nodes param
- Revert "[hcp] Filter regions where HostedCP is avalaible"
- removed --channel-group from --help options.
- Filter regions where HostedCP is avalaible
- HostedCP is avalaible
- update owners file
- STS is now default mode for cluster creation, added flags for non-sts
- Transformer added to change escaped empty strings to real empty strings
- refactor: manual aws command builder
- add
--yes
to create cluster cmd - fix hosted cluster parameter in create cluster
- unhide
--tags
during cluster create - Create/oidcprovider bug sets interactive.Enable
- fix: add tags check when b.tags nil
- Output current environment when it is not production
- Bump OCM SDK GO version to v0.1.292
- FIPS: Unhide flag
- Revert "[SDA-6643] STS is now default mode for cluster creation, added flags for mint mode/non-sts mode"
- feat: add warn messages about sts/non sts modes
- fix: specify which flag in message
- fix: message when non sts
- fix: message non sts
- fix: don't show if redirecting to file
- Bump OCM SDK GO version to v0.1.293
- show Limited Support status when calling
rosa describe cluster
- Create machinepool - filter supported instances by availability zones
- feat: add aws command builder unit tests
- Create a machine pool - prevent choosing a spot instance for a local AZ
- feat: new upgrade roles command and some refactors
- Fix bug - create a machine pool with a different region configured in the AWS CLI
- Create machine pool - display spinner when fetching instance types
- feat: add channel group and option to choose version for policy tags in upgrade roles cmd
- fix: sort prefixes to ensure consistancy when they are the same rank
- fix: order of messages
- fix: remove not needed vars in favor of using the args
- fix: prompt mode for upgrade cluster when sts and mode is empty
- fix: invert condition for no reason to update
- fix: always show warning, don't go into interactive if mint mode or non sts flags are enabled
- feat: checking undefined aws region
- fix: handle empty strings before validation
- Move hcp region supports check to the backend side
- Adding check for Changes in replicas flag too
- Remove redundant quotation
Release v1.2.8
- fix: path args need not to be explicitly set for interactive mode to ask about it
- fix: adding arn path validator to create account roles --path arg
- fix: clearer message
- fix: adding conditions for piping the output
- fix: setting path arg in a new line for all commands
- fix: differentiate between '/' and /
- fix: block managed services path option
- fix: remove error to add support for path in ARN
- feat: add validation to path ocm/user roles
- fix: consider empty path valid creating acc roles
- fix: accepts empty path
- fix: adding leading space before all path args when building commands
- fix: invert path detected message condition
- fix: aws empty path is different than ours
- refactor: less hacky
- fix: aws acc id on whoami
- fix: change message from one minute wait for several minutes
- [SDA-6984] Added unit tests
- chore: bump go ocm sdk v0.1.288
Release v1.2.7
- fix - Google IDP doesn't work when created with ROSA CLI
- Refactored ROSA to create operator policies when running rosa create cluster.
- move operator policies from account to operator commands
- SDA_4308: use root CA to generate OIDC thumbnail
- Support for ARN paths in iam roles and policies (ocm-role, user-role, account-roles, operator-roles)
- compare arn path for existing policy/role
- missing '--operator-roles-path' in 'To create this cluster again...'
- hide ARN path flags for some commands
- fix: missing changes for --role-path
- path for account and operator roles and policies
- fix: using installer instead of control plane role for path
- feat: unify account roles its policies paths
- feat: unify operator role and policy with path from account roles
- fix: getting path from master instance role
- feat: remove operator role path in create cluster in favor of master role path
- fix: remove operator-role-path from generated create cluster command as it was deprecated
- fix: remove path arg from -h ocm-roles description
- feat: adding message about operator roles and policies path
- Added redhatmanaged=true tag to operator roles in manual mode
- fix: enable path arg visibility
- fix: path compatibility issue with inline policies from account roles
- Create cluster - use a GET request to describe cluster details
- Refactor GetCluster function
- fix- It failed to set empty value with "" for no_proxy filed via interactive mode
- Add red-hat-managed: true tag to roles and policies
- red-hat-managed: true tag now added to operator roles
- Adding an info message after rosa delete admin
- bump ocm sdk to 0.1.285
- allow setting billing model for addon installations
- fix setting interactive mode enable for addon installation billing mode
- policies: Ensure policy version succeeds
- cluster: Allow using local AWS credentials
- Only display supported machine types by region
- Deleted account and operator policies
- ocm: Add aliases for local development
- Ensure prerequisites for deleting operator and account role policies
- fix manual create operator policy sda-6740
- Upgrade OCM-SDK-GO version to 0.1.287
- Added red-hat-managed: true tag to roles when rosa upgrade operator-roles is run
- Create cluster - list region filtered by OCP version
- Improve EnsurePolicy error message
- Add also local-proxy env config
- set mode only once in operator roles upgrade
- fix: changing description for channel group
- fix: description of version arg
- fix: reporting correct message back if specific version was chosen
- feat: adding -o yaml/json option to cmd whoami
- Upgrade cluster manual mode - prompt the aws operator role upgrade commands
- checking addon params
- fix: ':' character was at the wrong place
- Removing unnecessary check for managed services.
- Fetch all regions for non-interactive mode
- fix: adjusting order of calls to make sure deletion calls aren't being duplicate, this caused a 500 error on login after deleting and recreating admin from a newer rosa cli
- refactor: adding strategy and function to check if created on old ROSA
- Update stage console URL
- go: Bump version to 1.18
- Added red-hat-managed: true to manual operator/account/user roles creation
- lint: Remove deprecated linter
- fix: adding back ocm-roles path option and keeping it hidden
- feat: deprecate 'compute-nodes' args in favor of 'replicas' in create cluster cmd
- fix: adding trim spaces and tabs when validating cluster name
- chore: add gdbranco github user to owners file
- [SDA-5966]: Rosa STS mode auto conflicts with the watch option
- [SDA-6075] Add upgrade policy to rosa struct information when displayed with the rosa describe cluster with -o json or -o yaml
Release v1.2.6
- login: Remove token from error output
- debug: Remove AWS info from debug output
- add fake cluster parameter to create services
- fedramp: Update rosa-authenticator configuration
- network: Ensure there is no default network type
- Removed DisplayName from cluster
- Replaced display_name with name in query
- Removed change to query
- Create cluster - for single AZ, only allow to select one AZ
- Switch from github.com/pkg/errors to stdlib
- Updated SDK version and ran go mod vendor
- Ran go mod vendor after rebasing
Release v1.2.5
- Fix the typo in the error message when looking up the account role prefix
- fix for - Not able to remove or add a new cluster-admin in rosa cli fix for - Can't create a temporary admin user for ROSA cluster
- Create cluster - validate availability zones count interactively
- Delete admin should not delete htpasswd idp as the htpasswd list is not empty
- fedramp: Add environment-specific configuration (#702)