All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
1.86.0 - 2025-01-22
- Bump
cluster-standup-teardown
tov1.29.0
, which increases the timeout when creating clusters.
1.85.0 - 2025-01-21
- Ensure the storage test waits until PV has been fully deleted
- Attempt to delete any remaining PVs during
AfterSuite
before the cluster itself is removed to try to avoid leaving cloud volumes behind. - Show debug info if
Deployment
with private ECR image failed
- Updated
cluster-standup-teardown
to stop using spot instances for CAPA WCs
1.84.0 - 2025-01-17
- Refactored the storage test with the following changes:
- Split into individual it block for more fine-grained failures
- Added logging to show progress of tests while running
- Ensured the PersistentVolume is deleted before continuing to ensure we don't accidentally leave behind disks
1.83.1 - 2025-01-09
- Bump
clustertest
to v1.32.1 to fix timeout not being respected duringAfterSuite
.
1.83.0 - 2025-01-07
- Added a new test case to check for pods CrashLooping
- Re-enable
capv-on-capa
andcapv-on-capz
tests
1.82.0 - 2024-12-11
- Bump
cluster-standup-teardown
to v1.27.4 to use lower lifecycle hooks heartbeat timeout to allow spot instances to terminate more quickly (CAPA)
1.81.0 - 2024-11-29
- Attempt to get the owner team of any failing Apps and report them with test failures for use in notifications
- Set responsible teams for observability-bundle and security-bundle test cases
1.80.0 - 2024-11-25
- Introduced a new
SetResponsibleTeam
function to annotate test cases with the team responsible for their functionality passing.
1.79.0 - 2024-11-19
- Increase CAPVCD
ClusterReady
timeout to 40Min
1.78.0 - 2024-11-19
- Set CAPVCD
ClusterReady
timeout to 25Min in basic test
1.77.0 - 2024-11-19
- Allow to override
ClusterReady
timeout for cluster upgrade test - Set CAPVCD
ClusterReady
timeout to 25Min in upgrade test
1.76.4 - 2024-11-14
- Skip
capv-on-capa
andcapv-on-capz
tests until we can figure out VPN routing to the new vSphere provider - Bump
cluster-standup-teardown
to v1.27.3 to update CAPV cluster values for the new vSphere provider
1.76.3 - 2024-11-08
- Marked the EKS tests as skipped / pending until we re-focus on them again
1.76.2 - 2024-11-07
- Updated
cluster-standup-teardown
to 1.27.2 - Updated
cluster-test
to 1.30.2
1.76.1 - 2024-11-07
- Updated
cluster-standup-teardown
to 1.27.1 - Updated
cluster-test
to 1.30.1
1.76.0 - 2024-11-07
- Updated
cluster-standup-teardown
to 1.27.0 - Updated
cluster-test
to 1.30.0
1.75.1 - 2024-10-31
- Updated
cluster-standup-teardown
to 1.26.1
1.75.0 - 2024-10-22
- Updated
clustertest
andcluster-standup-teardown
to supportcluster-cloud-director
as a unified app
1.74.0 - 2024-10-15
- Debug the Cluster CR status when the workload cluster fails to standup correctly during
BeforeSuite
- Updated
clustertest
to include latest supported Release providers
1.73.0 - 2024-10-11
- Bumped
cluster-standup-teardown
to v1.25.7 to update the values used for CAPV and CAPVCD clusters. - Bumped
sigs.k8s.io/cluster-api
to v1.8.4 - Bumped
github.com/cert-manager/cert-manager
to v1.16.1
- Bumped
cluster-standup-teardown
to v1.25.5 to add proxy vars to CAPVCD test values - Bumped
clustertest
tov1.27.3
to include Provider fix when loading existing workload cluster
1.72.0 - 2024-10-08
- Added test to CAPA test suites to test pulling private images from ECR
- Extra failure handlers to some of the basic tests
1.71.2 - 2024-09-26
- Upgraded
cluster-standup-teardown
to add name property to the capvcd test values
1.71.1 - 2024-09-24
- Upgraded
cluster-standup-teardown
to remove node classes from capvcd test values
1.71.0 - 2024-09-23
- Moved common functions into
clustertest
to share between repos
1.70.1 - 2024-09-19
- Update
clustertest
with GitHub latest release fix
1.70.0 - 2024-09-17
- Upgraded Go to v1.23.1 and updated all modules to match
1.69.0 - 2024-09-06
- Use dedicated AWS Accounts for the different CAPA test suites (Private, EKS and "normal")
1.68.0 - 2024-09-06
- Updated clustertest with fix for version prefix on releases
- Updated all dependencies to latest versions
1.67.0 - 2024-08-27
- Updated
clustertest
with support for unified cluster-vsphere app.
1.66.1 - 2024-08-23
- Output some debug logging when Cluster fails to standup during
BeforeSuite
1.66.0 - 2024-08-22
- Added support for private CAPZ tests.
- Updated
cluster-standup-teardown
to latest to make use of private CAPZ cluster builder.
1.65.0 - 2024-08-19
- CAPV on CAPA tests.
- Updated
clustertest
andcluster-standup-teardown
to latest to make use of Teleport kubeconfig if available
1.64.0 - 2024-08-16
- Updated all modules to latest (including support for Kubernetes v1.31)
- Replace
containerdVolumeSizeGB
andkubeletVolumeSizeGB
withlibVolumeSizeGB
(from cluster-standup-teardown upgrade)
1.63.1 - 2024-08-08
- Upgraded
clustertest
to include fix for correctly handling Releases version prefixes.
1.63.0 - 2024-08-06
- Skip rolling update control plane tests if the control plane resource did not change
- Updated dependencies
clustertest
,cluster-standup-teardown
andteleport
to latest.
1.62.1 - 2024-07-26
- Bump cluster-standup-teardown and clustertest with latest releases SDK to correctly get latest release
1.62.0 - 2024-07-24
- Re-enabled
capa-private
tests because EC2 user data limit issue should be solved now that we store Ignition bootstrap data in S3 (issue)
1.61.1 - 2024-07-23
- Bump releases SDK to actually handle Azure
1.61.0 - 2024-07-23
- Bumped
clustertest
andcluster-standup-teardown
to support Releases with CAPZ
1.60.0 - 2024-07-22
- A framework for overriding default timeouts used by test cases. Introduces a new
timeout
internal package and new functions on thestate
that allows getting and setting a custom timeout per test suite.
- Update
cluster-standup-teardown
tov0.15.0
1.59.0 - 2024-07-09
- Updated
cluster-standup-teardown
to the latest release.
1.58.0 - 2024-07-08
- Added failure handling for Deployment, StatefulSet and DaemonSet not having expected number of ready replicas
- Updated
clustertest
to include logs when debugging failing pods
1.57.2 - 2024-07-05
- Switched to using
ShouldSkipUpgrade
fromclustertest
1.57.1 - 2024-07-05
- Allow upgrade tests to be run from Releases test pipeline
1.57.0 - 2024-07-05
- Switched to using DNS resolver and HTTP client from
clustertest
1.56.0 - 2024-07-02
- Updated upgrade tests so they can test upgrades to Releases
1.55.0 - 2024-06-28
- Made use of
GetWarningEventsForResource
fromclustertest
for Cert-Manager tests
- Use
failurehandler.AppIssues
to provide extra debugging when App-related tests fail (timeout)
1.54.1 - 2024-06-27
1.54.0 - 2024-06-27
- Update
clustertest
to v1.7.0 to support the new environment variable for controlling what Release to use when creating clusters
1.53.0 - 2024-06-25
- Added a test case to check the status of all Jobs in the WC
- Use actual types from Cert-Manager instead of using
unstructured
forCertificate
andClusterIssuer
- Refactored some functions to explicitly take in a context instance instead of using
context.Background()
so that timeouts set on the context can be respected
1.52.0 - 2024-06-25
- Updated all dependencies to latest version
1.51.0 - 2024-06-24
- Updated all depenedncies to latest version
1.50.0 - 2024-06-24
- Added a test case to ensure the hello-world ingress has a ready Certificate
- Included extra logging in ClusterIssuer test case to output the status and failing events of the post-install Helm Job.
1.49.0 - 2024-06-21
- Updated
clustertest
to latest with additional logging for node checks - Updated
cluster-standup-teardown
to latest with cluster-autoscaler config forscaleDownUnneededTime
1.48.0 - 2024-06-20
- Ensure timeout is always reset on the context in the AfterSuite to ensure enough time is given for cleanup
- Test if default apps are deployed before upgrading the cluster.
- Test workload cluster's Deployments, DaemonSets, StatefulSets and Pods before upgrading the cluster.
- Test if cluster is ready before upgrading the cluster (check Cluster resource Ready condition).
- Test if machine pools are ready and running before upgrading the cluster.
- Test if machine pools are ready and running in common tests.
- Update clustertest to v1.3.0 to support releases with cluster Apps for Azure.
- Updated teleport api module to latest available.
- Increase node pool min size from 2 to 3 for CAPA upgrade test.
- Disable spot instances for CAPA upgrade suite, as we suspect that using spot instances is causing Upgrade suite failures lately.
- Remove node checks from CAPA upgrade suite because we already check nodes in common tests after the upgrade.
- Additional logging added to the deployment scale test to display conditions of the pods and taints of the worker nodes
1.47.0 - 2024-06-14
- Add
china
test suite forcapa
provider. - Add wildcard ingress DNS check test to the
hello
test.
1.46.0 - 2024-06-13
- Update
clustertest
andcluster-standup-teardown
to latest releases - Switch to using new
ApplyBuiltCluster
in upgrade tests to avoid building the cluster twice
1.45.0 - 2024-06-10
- Update
clustertest
to v1.0.0 to support Releases with cluster Apps - Update
cluster-standup-teardown
to v1.5.0 to support Releases with cluster Apps
1.44.3 - 2024-06-10
- Update
cluster-standup-teardown
to v1.5.0
1.44.2 - 2024-06-06
- Update
cluster-standup-teardown
to v1.4.0
1.44.1 - 2024-05-17
- Update
cluster-standup-teardown
to v1.3.0
1.44.0 - 2024-05-16
- Add test for CAPA cluster in Cilium ENI mode
1.43.0 - 2024-05-16
- Wait and test for
security-bundle
apps to be deployed.
- Increase
kubeadmControlPlane
rollout timeout from 20min to 30min. - Increase node roll check time to 180 seconds from 25 seconds in the upgrade test.
- Reduce the timeout for default apps checks
- Skip checking default-apps-aws version when unified cluster-aws app is deployed.
1.42.0 - 2024-05-14
- Add support for unified cluster-aws app. With cluster-aws v0.76.0 and newer, default apps are deployed with cluster-aws and default-apps-aws app is not deployed anymore.
1.41.0 - 2024-05-13
- CAPV on CAPZ tests.
- Get base domain from Cluster values instead of default-apps values
1.40.0 - 2024-05-10
- Make error messages actionable for Prometheus metrics query test failures
- Disabled
capa-private
until we have fixed the userdata issue with MachinePools ongoat
1.39.1 - 2024-04-29
- Updated cluster-standup-teardown to v1.0.2
- Made use of new
clusterbuilder.KubeContext()
function - Updated readme requirements
- Removed KubeContext const from each test suite
1.39.0 - 2024-04-26
- Switched to using
cluster-standup-teardown
to handle cluster creation and deletion logic - Removed cluster and default-apps values in favour of defaults from
cluster-standup-teardown
1.38.0 - 2024-04-25
- CAPZ: Change Helm values structure where root-level properties are moved to
.Values.global
.
1.37.0 - 2024-04-19
- Add common basic test that checks if Cluster Ready condition has Status=True.
- Add upgrade test that checks if control plane rolling update has finished (if it has started).
1.36.0 - 2024-04-18
- Revert #169 by increasing CAPVCD CP nodes back to 3.
- Use default values of
cluster-cloud-director chart:v0.50.0
. - Update README example kubeconfig for CAPVCD to use
gerbil
asguppy
is dead.
1.35.0 - 2024-04-08
- Add support for mimir in the
metrics
test.
1.34.0 - 2024-04-04
- Include the MCs api endpoint when logging out the "checking connection to MC" message
- Switch CAPVCD to Flatcar (1.25.16).
1.33.0 - 2024-03-28
- Increase cluster creation timeout to 30 minutes.
- Switch CAPVCD values to 1.25.13.
1.32.0 - 2024-03-20
- Increase cert-manager test timeout to 2 minutes.
- Enabled EKS tests.
- Avoid checking control plane metrics in EKS.
1.31.0 - 2024-03-14
- Disabled EKS tests due to environment being broken
1.30.1 - 2024-03-14
- Set app versions when using
standup
CLI - Create the
results.json
file as soon as possible when runningstandup
in case the cluster creation fails to allows theteardown
command to still use it.
1.30.0 - 2024-03-12
- Add a new test that checks if key workload cluster metrics exist in prometheus.
1.29.0 - 2024-02-29
- Add teleport connectivity test
1.28.0 - 2024-02-29
- Bump Go version
v1.21
for teleport connectivity test - Update CAPV values and make more use of chart's default values
- Try and use ingress nginx app's config if present.
- Enable ingress test for private CAPA test.
- In environments where an egress proxy exists, don't use 8.8.8.8 as a DNS resolver for the ingress test as that would prevent the proxy address from being resolved.
1.27.0 - 2024-02-14
- Enable scale test for private CAPA clusters.
- Increase max node pool size to 5 for CAPA test suites.
1.26.2 - 2024-02-13
- Ensure upgrade tests install latest released version first
1.26.1 - 2024-02-12
- Use
Standard_D4s_v5
and 2 replicas for workers in CAPZ.
1.26.0 - 2024-02-08
- Add tests on DeploymentSets, StatefulSets and DaemonSets to ensure all desired replicas/pods are running.
1.25.4 - 2024-02-08
- Re-enable CAPVCD tests on gerbil.
1.25.3 - 2024-02-05
- Disable Bastion Support for CAPZ
1.25.2 - 2024-01-31
- Make PVC pod run unprivelleged.
1.25.1 - 2024-01-31
- Add security context to PVC pod.
1.25.0 - 2024-01-30
- Refactored
hello-world
tests to split out into individual (ordered) test cases to better understand where things fail - Increased the timout of the
hello-world
ingress check from 10min to 15min - Introduced a check for the ingress resources status being updated with the load balancer hostname
1.24.0 - 2024-01-26
- Set min worker nodes to 2 for CAPA clusters
- Make the scale test dynamic based on the current number of worker nodes
1.23.2 - 2024-01-26
- Use custom dialer when calling hello-world application to avoid DNS caching the not found result
1.23.1 - 2024-01-19
- Disabled CAPVCD tests while we recover the broken MC
1.23.0 - 2024-01-19
- Add a retrying healthcheck call to the MC at the start of the test suite to ensure connection is usable
- Re-enabled CAPA private test suite
1.22.0 - 2024-01-18
- Disable capa private test suite until MC stability issues are solved
1.21.3 - 2024-01-17
- Remove CPI config from CAPVCD tests so that we use the chart defaults instead.
1.21.2 - 2024-01-15
- Consolidated test suite setup (
BeforeSuite
andAfterSuite
) into a single, reusable module to reduce duplication.
- Added
AfterSuite
to perform test cluster cleanup. This should now cleanup clusters even if they timeout during creation in theBeforeSuite
.
- Added EKS cluster support to the
standup
CLI
1.21.1 - 2024-01-11
- Fix ClusterIssuer test to actually keep checking for issues being ready until timeout
- Fixed typo in CAPA private suite name
1.21.0 - 2024-01-11
- CAPA private cluster tests
- EKS autoscale tests
- Allow basic tests to be flakey and retry in case of network issues
1.20.4 - 2024-01-08
- Reduce the number of CP nodes from 3 to 1 in CAPVCD tests in order to avoid timeouts.
1.20.3 - 2023-12-19
- Add test config for upgrades to be able to customize timeouts per provider.
1.20.2 - 2023-12-13
- Disable Bastion tests for
capa
provider.
1.20.1 - 2023-12-13
- Update to new schema for
cluster-eks
values.
1.20.0 - 2023-12-11
- Bump
clustertest
to v0.14.0 that increased the char count of cluster names to 20 chars - CAPV: WCs have the default deny-all network policies
1.19.3 - 2023-12-05
- Increased timeout for observability-bundle apps.
1.19.2 - 2023-12-04
- Removed the old schema from the cluster-aws values
1.19.1 - 2023-12-04
- include both schema versions in CAPA cluster values file while migration is underway
1.19.0 - 2023-12-01
- Add initial support for new cluster app Helm values structure where root-level properties are moved to
.Values.global
.
1.18.0 - 2023-11-28
- Add
observability-bundle
test in the common/apps.go file.
1.17.1 - 2023-11-26
- Re-enabled CAPVCD tests
- Small cleanup tasks - added some relevant comments, explicitly set false bools, added logging to scale test
1.17.0 - 2023-11-15
- Add new test provider EKS.
- Add test for autoscaling MachinePool nodes.
1.16.2 - 2023-11-13
- Fix CAPV tests by switching to Flatcar image.
1.16.1 - 2023-11-03
- Updated
clustertest
to v0.12.1` to fix upgrade tests.
1.16.0 - 2023-11-02
- Updated
clustertest
to v0.12.0that now creates a ServiceAccount in the workload cluster to use for authentication when communicating with the WC in tests and updates
GetExpectedControlPlaneReplicasto handle managed clusters such as EKS (introduced in
v0.11.0`)
1.15.1 - 2023-11-02
1.15.0 - 2023-10-06
- Vsphere now uses IPAM also for WC's Service LB (kubevip)
- Moved
helpers
intointernal
directory - Moved
common
intointernal
directory - Removed unused function
- Cleaned up Readme
- Updated
clustertest
to v0.10.0 - Change control plane node test to use new
GetExpectedControlPlaneReplicas
helper function
1.14.0 - 2023-10-05
- Use Spot instances for CAPA tests.
- Use smaller volumes sizes for CAPA tests.
- Updated
clustertest
to v0.9.0. - Make timeouts more relevant to the context of the test
1.13.0 - 2023-09-15
- Updated
clustertest
to v0.7.0. - Refactored app deployment to use
clustertest
helper functions.
1.12.1 - 2023-09-12
- Refactored the various app status checks to make use of the helper functions in
clustertest
- Updated
clustertest
to v0.4.0
1.12.0 - 2023-09-12
- Add test that deploys hello-world application.
- Add a cert-manager ClusterIssuer readiness test.
1.11.1 - 2023-08-31
- Shrink container image by excluding build artifacts and language files.
- Rely on
cluster-aws
default values when installing the chart.
- Updated readme info
- Bumped
clustertest
to v0.3.1 with fix forClusterValues
unmarshalling if differentNodePool
types - Refactored node tests to reuse code now the
NodePool
type fix is inclustertest
1.11.0 - 2023-08-29
- Generate test result reports
1.10.0 - 2023-08-29
- Bumped clustertest to
v0.3.0
to be able to set the appOrganization
. - Bumped clustertest to
v0.2.0
to use custom app values. - Refactored default app tests so they can log out current state during iterations
- Log out the non-running pods in test spec to show progress
- Bump Ginkgo to v2.12.0
1.9.0 - 2023-08-21
- Add a PVC test.
- Bastion tests are disabled for CAPV.
- Refactored state management to use singleton
1.8.0 - 2023-08-14
- Add common test to make sure all default apps are marked as
deployed
. - Upgrade cluster tests
- Bumped clustertest to
v0.1.1
to fixGITHUB_TOKEN
issue. - Bumped clustertest to
v0.2.0
to use custom app values.
- Upgrade cluster tests
1.7.0 - 2023-07-20
- Abstract away the cluster app creation to a provider-specific function to handle provider specific requirements (such as VSphere credentials as extra config)
- Bump
clustertest
to v0.0.18
1.6.0 - 2023-07-20
standup
now supports specifying cluster app and default-apps app versions as arguments, defaulting to latest
1.5.0 - 2023-07-17
- CLI tool
standup
for creating test clusters outside of the test suite - CLI tool
teardown
for deleting test clusters previously created
1.4.0 - 2023-07-14
- Support marking test suites as skipped / ignored by prefixing the directory with
X
- Tests verifying control plane nodes and worker nodes are consistently Running
- Tests that verify cross-provider DNS records
- Marked CAPV standard test suite as active (removed the static IP for control plane -> relying on IPAM).
- Marked CAPV standard test suite as skipped until it can reliably run the tests.
- Changed parameters for CAPVCD tests by applying latest schema changes.
- Changed parameters for CAPVCD tests for
0.12.0
release.
1.3.0 - 2023-05-30
- CAPV standard test suite
1.2.0 - 2023-05-12
- Wait for Org to finish deleting when deleting cluster
- Increased Ginkgo suite timeout to 4 hour
- Continue running other remaining test suites when a suite fails
1.1.0 - 2023-04-19
- Updated readme with instructions on adding tests
- Ability to run tests against an existing WC. To do this export E2E_WC_NAME and E2E_WC_NAMESPACE
1.0.2 - 2023-03-31
- Prebuild the test suites in the Docker image to speed up starting the tests
1.0.1 - 2023-03-31
- Bumped custertest to v0.0.7
1.0.0 - 2023-03-30
- CAPA standard test suite
- CAPZ standard test suite
- CAPVCD standard test suite
- Example common tests
- Dockerfile for running tests in CI