jond

[yuvalk]

• Updates router metrics e2e
• extended/router: Omit routes.json from cat command
• test: add error output for build logs
• test: use ubi instead of centos
• Update OWNERS
• test/extended/prometheus: move checking telemetry to later steps
• test/extended: regenerate
• extended/apiserver: check for NonReadyRequests events for early traffic from LBs
• add a test to check for SCC denials during install
• Remove broken machine-api taint test
• don't wait for an OLM operator to succeed on non-amd64
• Bump to current openshift/kubernetes master
• Bug 1882454: test/apiserver: replace centos by ubi
• Bug 1861746: perform oauthaccesstokens only on upgraded clusters
• test: extended: add Late systemd timeout test
• Bump(github.com/openshift/kubernetes)
• make update
• refactor synthetic test creation from monitor
• add junit for illegal pod state transitions
• Update upgrade testmatching from [Suite:openshift] => not [Suite:k8s]
• test: add more debug info when we fail to wait for service account
• add test for failing network sandbox creation
• add test for systemd timeouts
• extended/apiserver: add test looking out for GracefulTerminationTimeout events
• Fix Cinder e2e tests not to leak volumes
• UPSTREAM: 94881: e2e: skip write/read cache when running Block-mode RWX PVC tests
• image_ecosystem: reinstate nodejs test with postgresql based app
• skip service type load balancer upgrade test on ovirt
• Add a new FIPS test
• hello-openshift: fix base images
• images/hello-openshift: Add OWNERS file
• examples/hello-openshift: Update OWNERS
• Awaiting for operators rollout to complete
• move jenkins e2e nfs setup from docker.io to quay.io image
• test: Increase allowed telemetry series by 100
• Fix openshift-sdn mode detection in e2e tests
• add event monitor test for kube-apiserver not being shutdown gracefully
• test: improve handling of pending pods crashloop detection test
• Make NewCLIWithoutNamespace not to use namespace actually
• Remove Tempalate, add WithToken and fix CLI usage w/o Args
• Squash all oc execute methods
• Move admin.sh to a proper e2e
• Remove unused testdata files
• Generated changes
• tests/extended/OWNERS: add ashcrow to approvers
• fail on pods moving back to pending in 4.7 and later
• hello-openshift: Echo listening port in HTTP resp
• use net/http instead of client-go for e2e PDB test
• handle cases where pod sandboxes fail after pod deletion
• tighten scc test to start failing on more than six pods failing
• bump(k8s.io/kubernetes)
• Fix up and regenerate test annotations
• Clarify the split between o/k and o/o e2e test annotations.
• Belatedly un-skip two passing tests
• Evaluate metrics of an e2e test.
• fix typo in word: strategy
• allow SCC to fail more often on fips
• move off docker.io in templates/builds/jenkins tests
• make update-bindata
• test/extended/router: Add OWNERS
• update SCC test to either pass or fail, but never flake
• Give informers time to observe limitrange
• check apiserver uptime with both new and re-used connections
• Squash methods into a single used one
• Move timeout.sh to a proper e2e
• Move debug.sh to a proper e2e
• Generated changes
• disable cakephp while cuppett/scl sort out new upstream regression
• make update
• Bug 1882556: Access git over https for tests
• move cakephp e2e skip down a level to make openshift-tests happy
• make update
• test "openshift.io/imagestreams" count quota
• generated
• add cakephp back in with upstream fix landing
• make update
• Update api references for etcd storage path data
• Export etcd storage data configuration for reuse by apply testing
• Ensure an etcd storage path stub that can pass validation for apply
• Ensure CLI.SetupProject returns the name of the project it creates
• Remove CSI driver installation from openshift-tests
• Change granularity of test monitor to 1s
• Monitor connectivity to API servers
• Add synthetic failures to stdout
• Start sampling immediately
• Add synthetic test duration
• bump(k8s.io/kubernetes)
• BZ 1838751: oVirt, enable Skipped network tests
• bump(*): Rebase to kube 1.19.2
• rename 'foo' build arg to avoid conflicts with 'foo' env's in the tools image
• make update output
• Add e2e testing of server-side apply for openshift types
• Update test annotations
• Bug 1896188: bump wait timeout for build completion in oc debug test
• Image extract tests should check results of ImageStreamImport
• Monitor availability of image registry during upgrades
• Make oc mirror tests aware of HTTP/2
• Updating ose-hello-openshift builder & base images to be consistent with ART Reconciling with https://github.com/openshift/ocp-build-data/tree/f82a216a6a3707b80a635bace9367f1a8288b7a7/images/ose-hello-openshift.yml
• Updating openshift-enterprise-tests builder & base images to be consistent with ART Reconciling with https://github.com/openshift/ocp-build-data/tree/f82a216a6a3707b80a635bace9367f1a8288b7a7/images/openshift-enterprise-tests.yml
• Reword CNV as OpenShift Virtualization
• e2e: node: filter out worker nodes without MCD
• Logging more during ephemeral registry deployment
• hello-openshift: Set BZ component in OWNERS
• Using test namespace to deploy ephemeral registry
• Using BeforeEach to set up ephemeral registry
• bump(*): Skip "subPath should be able to unmount" NFS test
• Monitor availability of registry through router
• hello-openshift: Make request-port header spec compliant
• [sig-cli][Feature:LegacyCommandTests][Disruptive][Serial] test-cmd: adjust tests to rebased 1.20-beta.2 oc command
• e2e: update explain tests with missing types
• e2e: add new test which looks for not covered types
• e2e: scrape CRD yaml when we fail regexp match
• e2e: add useraccessoauthtokens RBAC
• e2e: node: explicit image assignment for tests
• e2e/cli: drop rsync.sh which is already well covered in rsync.go
• e2e/cli: move policy-storage-admin.sh to proper e2e
• Generated changes
• test: Enable mirroring test images to a different repository
• test: Add a framework for mirroring origin images
• e2e/cli: add missing GinkgoRecover
• e2e/cli: move observe.sh to proper e2e
• e2e/cli: add oc observe CRD test
• Generated changes
• Update etcd data locations for kube 1.20.0-beta.2
• test: Update openshift tests to use locked images
• [sig-auth][Feature:LDAP][Serial] ldap group sync can sync groups from ldap: skip the test, it's broken
• Replace oc --config with oc --kubeconfig
• Fix cluster quota test to tolerate a new default config map in every namespace
• Tolerate 1.20 error message in test-cmd
• Update generated
• Temporary disable tests broken upstream
• Revert "Temporary disable tests broken upstream"
• Fix panic dereferencing nil
• clean up remaining build busybox refs after claytons overhaul
• make update output
• Do not set status for DeploymentConfigs
• Fix field names for pods and routes
• Fix setting dockerImageLayers for test images
• Get rid of duplicates from test/integration/testdata
• Generated changes
• Skip mco CRD checks for ibmcloud provider
• Bug 1904679: drop docker.io from image tests
• test: Append image test should be relative to the tools image
• add kubevirt as a provide to openshift tests
• allow operators to settle after MCPs are done upgrading
• Add consolequickstarts to list of console resource exceptions
• Add libvirt to unsupported platforms for LB service
• normalize ginkgo structuring so we avoid timing windows changing users with k8s e2e setup logic (and hope memory is OK)
• make update output
• extended-tests: kube-apiserver: switch flowcontrol storage to alpha
• Fix build test CM count
• tolerate apiserver restart errors when creating machines in quorum restore test
• increase the wait time for machines to re-create
• Improve reliability of ssh command execution in disruptive testing
• Skip tests incompatible with disruptive testing
• Remove obsolete DR testing of snapshot restore
• Temporarily skip machine recover DR test pending a fix
• Ensure quorum restore DR test can take a backup when api is down
• make update
• Skip table tests broken by 1.20 rebase
• make update
• test: Update images README and add OWNERS
• add build s2i as root uid image for associated tests
• make update output
• update s2i_root.go to use CI image that will be mirrored to quay.io/community-images
• Bug 1896877: Enhance host name validation
• Bump(*): o/k 1.20.0
• Revert "Skip table tests broken by 1.20 rebase"
• React to changes
• Disable broken tests
• Update etcd storage test
• Don't run disabled tests
• Update generated
• Renable fsgroup policy test
• Add openshift/network/third-party suite, for CNI plugin conformance
• test: Assess average series rather than max over the test window
• test: Allow tests that check invariants over time to be constrained
• server-side apply e2e: fix oauthclientauthorization race with KCM
• Replacing centos with base test image in image-annotation test.
• test: Fail if nodes go unready during a normal parallel run
• test: Only consider rule evaluation failures in current window
• test: Clamp metrics queries by default to the suite execution
• Bump(github.com/openshift/kubernetes)
• Add consoleplugins to list of console resource exceptions
• Updating ose-hello-openshift builder & base images to be consistent with ART Reconciling with https://github.com/openshift/ocp-build-data/tree/b0ab44b419faae6b18e639e780a1fa50a1df8521/images/ose-hello-openshift.yml
• Updating openshift-enterprise-tests builder & base images to be consistent with ART Reconciling with https://github.com/openshift/ocp-build-data/tree/f01352e0069d4e8cedc0880ae427359a96668512/images/openshift-enterprise-tests.yml
• broaden acceptance of scl sample s2i image usage messages
• Bug 1914894: make cmd tests more resilient
• replace test/cmd/builds.sh direct/indirect docker.io refs with imagestream image refs
• make update output
• move ldap group sync testdata to bindata
• move ldap groupsync tests from bash into golang
• regen bindata
• go mod tidy && go mod vendor
• remove ldap groupsync test from the broken set of tests
• Bug 1896188: wait for RC presence before starting debug session
• Bug 1914894: make tests more resilient
• UPSTREAM: 96990: Increase preemption timeout from 1 minute to 2 minutes
• tests: Synthetic tests moved up and out of test infra
• test/ginkgo: Stop printing inline events
• test/ginkgo: When --count is used tests all fail together
• test/ginkgo: Improve output when multiple tests are being run
• add additional settling debug information
• UPSTREAM: 94684: e2e: Pod should avoid nodes that have avoidPod annotation: clean remaining pods
• Bug 1896188: bump build timeout to 10m and deploy to 2m in oc debug
• Clean and replace docker.io images with other registries in examples
• Remove RHEL image streams from examples
• Bug 1914894: get stdout only when invoking oc process
• Bug 1904684: Update tests to use refreshed image-streams file and get rid of docker.io
• Generated changes
• add test to verify build /run fs contents, verify /run writeable
• make update output
• Bug 1921088: don't wait for PV removal in volumes.sh
• Generated changes
• bug 1908677: reenable SCTP Host Port test
• test: etcd leader changes should be measured over the test, not the run
• test: Router should require multiple passes before continuing
• Bug 1904684: fix missing bits in images.sh tests
• Additional fixes after docker.io changes to printer.sh and builds.sh
• Add e2e test for liveness override annotation
• test/cmd/images.sh: create imagestreamtag --from is a tag
• idle.sh: Drop custom endpoints from idle tests
• test: Use wildfly 20 since 19 can no longer be accessed at quay
• allow OKD /run contents for build fs test
• test: Ensure early|late tests are run with other suites
• test: SCC wait loop shouldn't be using 15s pauses
• test: Use a simpler namespace for tests that must create an execpod
• test: Don't initialize the cloud provider in run command
• test: Add --fail-fast to the tests command
• test: Inline custom suite logic into the suite definition
• test: Purge use of klog v1
• vendor: Remove klog v1 fully
• test: The disruptive suite should not have to skip certain lates
• test: The node ready test should require the node be up 6m
• test: Upgrade should prereq early tests
• test/extended/router: add e2e for idle/unidle
• Update bindata
• Bump(github.com/openshift/kubernetes)
• Allow RHE7 /run contents for build fs test
• tests: Upgrade tests should always include success output
• tests: Double up the network stress test pressure
• Fix CSI snapshotter image version
• Change registry.svc.ci.openshift.org to registry.ci.openshift.org
• vendor: Update to latest Kubernetes for test fixes
• vendor: Pick up ginkgo v4.7.0-origin.0
• test: Bypass upstream suite before/after because they are not useful
• test: Set cloud zone overrides in our test suite based on machines
• test: Bypass all upstream test waits for things we don't support
• test: Set provider if --provider passed with run -f -
• dr: Add test of documented backup and restore procedure
• dr: Skip quorum restore test pending the fix in Ensure quorum restore test follows the documented procedure openshift/origin#25774
• test: Set provider in all suites or explicitly choose skeleton
• Update test annotations
• Updating ose-hello-openshift builder & base images to be consistent with ART Reconciling with https://github.com/openshift/ocp-build-data/tree/5a1293dd0f380abf50c12d65c36655486d7745d0/images/ose-hello-openshift.yml
• Updating openshift-enterprise-tests builder & base images to be consistent with ART Reconciling with https://github.com/openshift/ocp-build-data/tree/5a1293dd0f380abf50c12d65c36655486d7745d0/images/openshift-enterprise-tests.yml
• bump(openshift/kubernetes): balance nodes in scheduling e2e
• make update
• Upgrade openshift/api to 670ac3fc997c
• test: Allow prometheus state clear to pass etcd leader change test
• Bump(github.com/openshift/kubernetes)
• Regenerate test annotations
• Test that new single replica topology API is taken into account
• adjust route host for redis/nodejs to accomodate new apiserver restrictions
• duplicate test/extended/builds/OWNERS in test/extended/image_ecosystem
• e2e: test OAuth API connections in the tests by that name
• test/extended/router/idle: address flakes/failures seen in CI
• tolerate permissions for allowing in-cluster CA bundles to all authenticated
• test: Calculate leader elections over individual metrics
• test: add vsphere to unsupported platforms for LB service
• test: Synthetic failures were not failing the job
• Added openshift-image-registry to single node topology test allowedToFail list
• test: AWS and Azure should tolerate no API disruption
• test: SCC early failures (before suite) flake until fixed
• test: Ensure nodes stay ready during upgrades after reboot
• Bug 1921720: fix sig-cli flakes
• Revert "test: SCC early failures (before suite) flake until fixed"
• refine SCC check to take into account the pod delay before failure
• Bug 1895093: Verify builds can mount proxy trustedCA
• Restore s2i memoryswap quota test
• test: Prevent all SCC failures from being allowed during normal e2e
• monitor: Improve cluster operator transition message
• test: Detect when the master pool is still updating after upgrade
• test: Don't allow alerts to fire during a test run
• Bug 1928297: Wait until router pod is ready before checking health
• Bug 1933114: test/extended/router/idle: skip test on virt platforms
• test: Exclude AggregatedAPIDown alert caused by test cleanup
• Remove console deployments from list of 2 replicas in single-replica topology
• Skip tests that rely on Machine API when Machine API is not installed
• Ensure that machine CRDs are installed before running Machine API tests
• test: Graceful termination invariant should separate pre-suite events
• test: Nodes that are deleted should not fire the unready alert
• test: Consider UpgradeRollback as an invariant reset for node ready
• Don't use atomic: in test/extended/images/signatures.go
• Re-enable test/extended/images/signatures.go
• flake on operator state changes during e2e
• test: Node unready should ignore the ns/ in event locator
• Add BareMetalPlatformType into e2e upgrade service unsupported list
• test/extended/etcd: fix leader change test
• split operator condition tests into separate tests for easy assignment and tracking
• operators: prevent panic in no crashlooping pods in core namespaces
• straight moves of the various synthentic tests
• DNS: Add new dual-stack service DNS test
• add monitor events for running tests
• indicate which tests correspond to an operator outage
• test: use Ruby 2.7
• test: When node upgrade sythetic resets, clear intermediate state
• test/extended/testdata/cmd/test/cmd/set-image.sh: use 2.7-ubi8
• test: Fail if a daemonset has a different upgrade strategy than default
• Add new tests for baremetal platform
• test: Prometheus helper should use thanos-querier
• Skip maxUnavailable check on DS on masters with complex label selectors
• test: Move apiserver graceful failure responsibility to invariant
• test: Temporarily move some synthetics out of disruptive
• e2e CLI: create SHAed tokens for tests
• etcd storage: prefix oauthaccess/authorize tokens with sha256~
• test: Correctly detect non-mirrored images
• test/extended/operators/daemon_set: Multus bug has been fixed
• test: Fix regressions in missing images
• Bug 1950095: Verify layer MIME types in built images
• test: Prevent limits from being set on pods and enforce requests
• test: Mark pod transition back to pending as a flake for now
• test/e2e/upgrade: Relax 'too long' soft timeout for rollback jobs
• test: Require no alerts during upgrades
• test: Grant bare metal components an exception for not having requests
• pkg/synthetictests/operators: Include the reason string in testOperatorState
• upgrade/upgrade.go: Enhance upgrade ack time out error
• test: stop building URL's with host/port and Sprintf
• replace registry.redhat.io refs with internal image registry refs in image-eco e2e
• make-update-output
• test: Mark kubelet terminates non-gracefully as a flake
• test: Synthetics should only return junit tests instead of a bool
• test: grant ovirt-csi-driver an exception for not having requests
• test: App availability test during upgrade must set minReady
• test: add baremetal provider
• remove unused initiatedAt in favor of a single At for monitor events
• formalize monitor event notification
• Events actually returns EventIntervals. Update method name to match reality
• cmd: initialize baremetal provider
• test: remove apk add tcpdump to internal ports test
• Updating openshift-enterprise-tests builder & base images to be consistent with ART Reconciling with https://github.com/openshift/ocp-build-data/tree/e3b61ff1dde2bfc2bc7f2b5efdd155d1d3299cd7/images/openshift-enterprise-tests.yml
• Sync perl/dancer imagestreams and use Perl 5.26-ubi8
• test: skip checking nodejs 10 and perl 5.26
• test/extended/single_node: update for monitoring components
• move the monitor API to an API
• allow separating monitor event identification from monitor event generation
• tests/extended: Add a basic sanity test for CoreOS bootimages
• Restore private git clone tests
• test: Do not check github for commits on OLM test
• test: Remove exceptions in fixed bugs in alert/resource tests
• test: Mark GCP as fixed and AWS as flaky in upgrade availability
• test: EventLevel is compared by value for filtering
• Introduce experimental minimal/reliable test suites
• test/extended/operators: Multus grew requests, drop its exception
• Fix tests to allow landing k8s bump in oc
• fix remaining token format discrepancies in parallel tests
• Remove the test for the authenticator accepting old-format tokens
• test: grant exception to installer etcd retry pods
• require operator availability and non-degraded for upgrades
• test: provide a way to skip tests on disconnected clusters
• Update test rules to bump k8s
• Change BindNetwork of outhserver to tcp to support IPv6
• upgrade: allow upgrades without version change
• test/extended/router: add certgen
• test/extended/router: add shard helpers
• test/extended/router/http2: delete obsolete files
• Baremetal platform tests refinement
• create operator intervals from events
• switch operators to use interval
• update e2e tests to use intervals
• serialize the event monitor events
• visualize monitor event intervals
• Add IPv6 check to support observe parallel tests
• test/extended/operators: remove CPU exception for Thanos querier
• pkg/test/ginkgo/cmd_runsuite: Fix e2e-* clobber for upgrade-conformance
• SNO topology: the oauth-server should only have a single replica now
• e2e requestheader: properly wait for a new oauth-server rollout
• test/extended/router/http2: rework and enable test
• test/extended/router/grpc-interop: delete obsolete files
• test/extended/router/grpc-interop: rework and enable test
• test/extended/router: change getHostnameForRoute signature
• Remove NAME parameter from shard template
• test/extended/testdata: regenerate bindata
• test/extended/router/h2spec: rework and re-enable test
• test/extended/testdata: regenerate bindata
• test/extended/operators: Require Upgradeable=True or unset for "start all core operators"
• update list of canary tests
• Fix oauthaccesstoken test to use the new required name
• test: replace dns test with known image
• upgrade: add "none" suite which skip disruption tests
• Add unit tests for "openshift-tests --provider"
• Rename test/extended/util/cloud to test/extended/util/cluster
• Move test skip/match handling into ClusterConfig
• Remove a weird short-circuit case in test config detection
• Simplify JSON --provider handling
• Clean up network plugin detection/config
• Auto-skip IPv6-only/dual-stack-only/SCTP tests based on network config
• test/extended/router: skip h2 related tests on OpenStack
• Remove non-conformance tests from minimal suite
• test: Remove legacy events that are no longer used
• test: Expose node changes as intervals
• monitor: Remove pointers to events where possible
• monitor: Record events at the time they occur
• monitor: Set a minimum width on the chart for smaller devices
• Makefile: Correct typo in make syntax for wildcard
• ensure that sampled conditions always have a range of time in their event interval
• upgrade: register none suite
• Bug 1949928: Remove references to EOL Image tags
• test/extended/router: h2spec increase timeout and retry tests
• test/extended/operators/resources: Marketplace Job sets requests
• cmd: remove reference to EOL perl image
• test/extended/router: skip http2 test when proxy is configured
• test/extended/router: Follow up to PR 26083
• test/extended/operators/resources: Autoscaler sets requests
• Monitor pod displacement during upgrade
• monitor: Add pull time to pulled events for analysis
• e2echart: Handle intervals that extend past the test range
• monitor: Clamp intervals to the test duration
• test/extended/router: skip http2, grpc-interop and http2 on AWS
• Bug 1848408: fix validation issues in test files
• monitor: mark the starting point for clusteroperators
• request to mirror nfs server image used in jenkins tests
• Bug 1949050: fix images.sh script
• allow HighOverallControlPlaneCPU during e2e runs
• update openshift/api and openshift/client-go for apirequestcounts
• add e2e test to block usage of removed APIs
• allow high CPU alerts to be firing and pending
• test: Mark known alert failure for auth operator to bug
• test/extended/operators/resources: Packageserver already sets requests
• test: Disable upgrade disruption check on GCP
• test: Alert check should check all conditions
• test: Allow KubeAPIErrorBudgetBurn on GCP platforms only
• monitor: Move "crashlooping pods" test to a synthetic
• test/extended/router: Fix-up Unidling test
• Bug 1949991: Ignore Catalog update pods in openshift/conformance/parallel
• test: Wait for service account token
• test: Report FailedCreatePodSandBox for multus -> API as known issue
• test: Exclude terminal pods from ForceDeletion
• [test/oc_tag] Add internal option for registry info
• Components should have required Priorityclass
• openshift-tests: fix a --provider --dry-run edge case
• Skip ClientSecretWithPlus test that uses oauth-openshift route for ibm clusters
• test: Associate ClusterOperatorDown for machine-config with bug
• Update node-states section in e2echart with roles info
• test: Add the Cinder client to the test image
• test: Suppress the SCC failure when a limit is present
• test/extended/router/idle: Only run on OVNKubernetes or OpenShiftSDN
• fix test README link
• must-gather: add oauth-apiserver to expected audit logs
• bump(k8s.io/*): 1.21
• React to bump
• Add custom timeouts to e2e framework
• Bug 1945085: enable back etcd test
• Update images based on k8s bump
• Add exceptions list for test images
• Remove unused test files
• Disable networking hostNetwork test
• Disable crictl test
• Disable firewall rule test
• Disable Dynamic Provisioning Invalid AWS KMS key test
• Disable sig-network tests failing on e2e-metal-ipi-ovn-ipv6
• Disable GlusterFS since it's not supported in OCP 4.x
• Disable ttlafterfinished job test
• Disable pull from private registry with secret test
• Generated changes
• audit-logs: allow .lock file and flake on lock.log
• audit-logs: re-add remove must-gather RemoveAll call
• dr: Remove unused fns getPullSecret and GetImagePullSpecFromRelease
• dr: Get quorum restore passing
• must-gather: check file name instead of whole path
• Router: Switch from networking v1beta1 to v1
• monitor: Allow samplers to be edge triggered by recording intervals
• Enable SNO check for image registry
• test: Report FailedCreatePodSandBox for multus -> API as known issue (v2)
• Bug 1959564: builds: skip /run verification tests
• test: Record disruption events in a way that shows up in the log
• test: SetupNamespace in cli helper should wait for default service account
• test: Wait longer for namespace deletion in a few cases
• test: Add endpoint availability events to chart
• Bug 1961057: Add [Driver: gluster] tests into unsupported list
• DNS: Use endpointslice v1 for DNS dual-stack test
• move verify run test from tools to cli image to avoid additional /run content
• Do not install python3-cinderclient on s390x
• BUILD-250 updating roots2i image for its multiarch analogue
• Add statefulsets to single-replica topology test
• Remove duplicate entry for Gluster tests
• Bump to get latest k8s deps
• vendor files
• React to bump
• Update test rules again
• test/extended/pods/priorityclasses: Improve error actionability
• Bug 1929396: Revert part of PR Bug 1828858: use LoadBalancers with ExternalTrafficPolicy set to Local openshift/origin#25406 - use ExternalTrafficPolicyLocal
• Remove network policy tests
• bump(k8s.io/*): 1.21.1
• Update images based on k8s bump
• test: Tolerate no AWS disruption during control plane upgrade
• narrow verify /run test to /run/secrets
• Disable 'Networking should provide Internet connection'
• Generated changes
• test/extended/operators: Drop service-ca Upgradeable exception
• Skip cluster quota test while bound token projected volume is being re-enabled
• test/extended/cli/mustgather: Local-only getPluginOutputDir
• Bug 1964334: use Bearer token to access prometheus /metrics endpoint
• Bump k8s
• Revert "test/extended/operators: Drop service-ca Upgradeable exception"
• BUILD-250 updating simples2i image for its multiarch analogue
• Dockerfile.rhel: fix a bash typo
• resources: remove metal3 from broken pods
• bump(openshift/kubernetes)
• Update README to include '410 Gone' workaround when updating vendoring
• [sig-arch][Late] operators should not create watch channels very often [Suite:openshift/conformance/parallel]
• test: Tolerate control plane unavailability before 4.8
• Bug 1968558: properly remove project after test ends
• Use cluster-logging for operator installation smoke test.
• Add test/extended/operators OWNERS file
• Add new E2E test for [SDN-1364] ACL audit logging
• Add OCP SDN Plugin checker
• test/extended/etcd/leader_changes: Explicit 'expected 0'
• test: Don't reuse route.spec.host in tests run in parallel
• test/e2e/upgrade/alert: Allow AggregatedAPIDown to unblock 4.7->4.8 CI
• Re-enable cluster quota test
• upgrade test: expect upgrades to take longer on AWS
• clarify failure message
• uprgade: extend upgrade duration to 105mins on AWS
• undisable NetworkPolicy tests
• UPSTREAM: : Replace cadvisor for 2868 upstream patch
• builds: update nodejs in pullsecret test
• Drop RHEL imagestreams from update-external-examples.sh
• Drop mongodb from external examples
• Update external examples
• cluster: sync quickstarts
• cluster: switch from nodejs-mongodb to nodejs-postgresql
• deployments: switch deployment-example from mongodb to postgresql
• builds: move multi-namespace-template from mongodb to mariadb
• jenkins: move pipeline tests from mongodb to mariadb
• image_ecosystem: do not wait for mongodb imagestream
• builds: use UBI8 ImageStreamTags
• test: Use last_over_time in pendingAlertQuery
• pkg/synthetictests/networking: Grant temporary exception for "error getting pod: Unauthorized"
• test: Frontends tolerate no disruption
• Updating openshift-enterprise-tests images to be consistent with ART Reconciling with https://github.com/openshift/ocp-build-data/tree/691e628254f318ce56efda5edc7448ec743c37b8/images/openshift-enterprise-tests.yml
• Add failing tests in OVN-K into skipped list
• test/extended/prometheus/OWNERS: refresh members
• builds: use ubi8/ruby-27 instead of centos/ruby-27-centos7
• builds: comment out multi-namespace template test
• cli: use ubi8/ruby-27 instead of centos/ruby-27-centos7
• cmd: use UBI8 ImageStreamTags
• cmd: use ubi8/ruby-27 instead of centos/ruby-27-centos7
• imageapis: use UBI8 ImageStreamTags
• image_ecosystem: use UBI8 ImageStreamTags
• jenkins: use UBI8 ImageStreamTags
• Update bindata
• Remove shawn from extended test owners (he moved to a new role)
• test/extended: add etcd OWNERS
• test/extended: Ensure all containers in the OLM deployment resources contain the IfNotPresent image pull policy
• Add renamed Prometheus test into skipped disconnected list
• test/e2e/upgrade/alert: Allow some Watchdog changes
• testPodSandboxCreation: skip sandbox errors for pods which were not deleted during network update
• Bump(github.com/openshift/kubernetes)
• Regenerate annotations
• change apiextensions to v1
• move jenkins e2e's that were using mongodb, then mariadb, to postgresql, to get them to pass again
• test/extended/prometheus: add an upgrade test for prometheus
• builds: comment out RHEL-only sample-pipeline template test
• remove apiextensions.k8s.io/v1beta1 and admissionregistration.k8s.io/v1beta1 from etcd tests
• Bug 1977454: Use nodejs to test service connection
• Increase OVN upgrade timeout by 15m
• Update quickstart examples and re-run make update
• test/e2e/upgrade: Add manifest delete annotation test
• vendor update
• Bug 1957886: re-enable TTLAfterFinished test job
• Skip Prometheus upgrade test if persistance storage is not enabled
• Fix make update-examples
• always discard v1beta1 in etcd test of apiextensions.k8s.io and
• Skip new NetworkPolicy tests for Network Stress
• test/e2e: allow workload incorrectly spread alert
• comment out broken local git, env var, jenkins pipeline test (moving to remote repo and openshift/jenkins)
• csrapprover: fix deprecated CSR api client
• on SNO, skip distruptive or multinodes tests
• Bug 1977097: Improve GC Check for Builds
• Bug 1942164: Fix time calc ordering for upgrades
• Enable TestEndpointAdmission test only for OpenShift SDN
• test/e2e/upgrade/upgrade: Unify duration-overshoot test-case name
• disruptive: output millis when disruption is < 1s
• upgrade: fix typo in upgrade filter for connection checks
• bump(*)
• adding tests for build volumes
• jenkins multi namespace test has been moved to client plugin tweak jenkins imagestreamtag podtemplate test to label imagestream
• test/extended: Remove the PackageServer deployment from the single-node
• disruption: tolerate on gcp for frontend
• oauth-server certs e2e: add RBAC rule to be able to read the distributed cert in a CM
• disruption: allow on OVN during upgrade for ingress
• pkg/synthetictests/operators: Set duration for operator-state transitions
• Remove auto trigger
• test/e2e/upgrade/alert: Temporarily allow HighOverallControlPlaneCPU
• Disable tests for k8s 1.22
• Generated changes
• initializeTestFramework: initialize TextContext ip family
• Ignore upgradable false condition on TechPreview clusters
• account for both cgroupv1 and cgroupv2 in s2i quota test
• sort upper bounds by operator name
• Bump upper bounds based on the latest 4.9 measurements
• Do not allow an excessive count of an individual event: they indicate a failure
• Disable tests for k8s 1.22
• Generated changes
• Add tests required internet into Skipped:Disconnected list
• Copy paste aws upper bounds for single node
• Set single node upper bounds
• Update canary/minimal e2e suite for 4.9
• Re-disable the IPV6 should provide Internet connection test
• temporary search for memory.max in build container during quota test
• test/e2e/upgrade/alert: Temporarily allow ExtremelyHighIndividualControlPlaneCPU
• Ignore openshift-marketplace pods
• OWNERS: Declare Bugzilla component
• Disable tests for k8s 1.22
• Generated changes
• Bug 1988828: fix test - reduce minimum size of expected files in must-gather tests
• Temporarily disable oc observe single test
• Skip the "Netpol" suite
• refine apiserver downtime sampling and calculations
• Revert "disruptive: output millis when disruption is < 1s"
• separate allowed upgrade duplicate events from generally allowed duplicate events
• pkg/synthetictests/operators: Drop Progressing from "stable" transition checks
• expose rest.Config from monitor
• pass a rest.Config to invariant checking and allow console readiness failures during install
• add several skip cases for pathological events
• tidy up the duplicate event handling
• kubectl: mark exit codes test flaky until 1.22 rebase
• Bug 1990988: Updating s2i images for php
• Added E2E test for multus-cni
• Skipped non-standard openshift installed crd testing for ibmcloud
• e2e/cli: move half of builds.sh to proper e2e
• generated changes
• Exclude local volume conflicts
• disable cli oc builds tests on disconnected clusters
• Allow TechPreviewNoUpgrade alerts when running on a TechPreview cluster
• Remove kubectl command exit codes test from canary
• search for correct memory.max file with cgroupv2 and build quota test
• Bug 1969471: HAProxy tests in sdn-network-stress job are flaky
• Add proxy support to OAuthServer tests
• bump(k8s.io/*): v0.22.0-rc.0
• Change imports for go.etcd.io/etcd v3.5.0
• React to k8s, grpc and library-go bumps
• Get rid of --viper-config which was removed in k8s
• Bug 1986001: bring back api data and resorucequota tests
• Update test images based on k8s bump
• Update etcd data tests based on k8s bump
• Retry kubeconfig checks, when kube-apiserver is temporarily unavailable
• account for cgroup v2 swap file only having swap for its amount, while cgroup v1 has memory+swap for its amount
• Add me to test approvers so I can approve annotation changes
• Add 'DeploymentAwaitingCancellation' as an allowed repeated event
• Disable [sig-network] Networking should provider Internet connection for containers using DNS
• Generated changes
• Switch kubeconfig retries to use backoff
• duplicated events: update list of known issues
• test: skip tests that won't work behind a proxy automatically
• Enable CSI hostpath tests
• Enable back HAProxy idle service connect test
• Skipped renamed dualstack tests in OVNKubernetes
• cli: use ubi8/ruby-27
• BUILD-188 image trigger tests update for LastTriggeredImageID
• Fix regex pattern for DeploymentAwaitingCancellation event
• kubelet: detect process overlap during apiserver rollout
• Bug 1995198: Skip admin can subscribe to operator using olm test on arm64
• e2e-oauth:requestheaders: log body of a failed authn request
• add API rollback resiliency test
• CustomNoUpgrade should not fire no-upgrade failures
• Bug 1986307: Bump openshift/kubernetes to re-enable tests
• Disable service internal traffic policy network tests for ovn-k
• Generate test rules list following bump of k8
• Re-enable test for running crictl on a node
• Bug 1986562: check buildConfig spec LastTriggeredImageID is empty
• Add imagepull backoff to frequent event allowlist
• operators: properly skip OLM test on arm64
• bump(k8s.io/*): v0.22.1
• Generated changes
• add special message and documentation of disruption time validation
• Bug 1989505: enable back oc observe test
• test/extended/operators: Remove Evan's GH handle from OWNERS
• Put both "no external IPv6 connectivity in CI" skips together
• Move broken dual-stack podCIDR test to [Disabled:Broken]
• rewrite idling tests to not be [Local]
• pkg/synthetictests: add etcd quorum-gaurd duplicate events to known issues
• Skip two unidling tests on ovn-kubernetes for now
• Bug 1989417: bring back oc adm storage-admin test
• kubectl: re-enable exit codes test
• Revert "Bug 1952457: Re-enable crictl node test"
• Split the "Cluster frontend ingress" test into 4
• watches: increase upper bound for marketplace operator
• Run k8s test separately from OpenShift tests
• Bug 1997509: delete only necessary resources instead of all
• CustomNoUpgrade should not fire no-upgrade failures
• must-gather: Ignore startup logs in kube-apiserver audit logs
• test/extended[auth*,oauth]: configure OWNERS
• images: port image signature workflow test to OCP4/UBI8
• builds: fix BUILD_LOGLEVEL tests for nightlies
• track the final state of pods and events from e2e tests for later debugging
• tests/extended/builds: handle new step logging
• bump(k8s.io/*): v1.22.1
• disable flaky unidling test
• Add [Serial:Self] label for tests that can't be run in parallel with a copy of itself Exclude these tests from "openshift/network/stress" suite
• Bug 1992596: e2e/cli: Move annotations.sh and basicresources.sh to proper e2e
• pkg/synthetictests: add etcd client connection closing to known events
• builds: skip private git repo tests outside of CI
• Use PlatformStatus Type instead deprecated Platform
• Refactor RestrictedEndpointsAdmission test
• Revert "Bug 1960674: images: port image signature workflow test to OCP4/UBI8"
• test/extended/prometheus: Validate alerting rules
• Remove exception for ovirt resource requests
• Bug 1999561: Extend API rollout resiliency test to validate pods restart during the rollout
• *: update transport package to 3.5
• vendor: tidy
• etcd-storage-test: mark node ready to save it for 5min from cloud provider deletion
• Fixup request/limits test to accomodate Pods annotated for workload partitioning
• add disable-scc label to kube e2e tests
• Make RestrictedEndpointAdmission test check EndpointSlices too
• stop adding individual serviceaccount namespace groups to SCC for e2e tests
• vendor: e2e iperf2 change threshold to 10MBps
• This would make CI for 4.10 in vSphere environment wee bit happier
• record only known images test pass so we can get a percentage of passes
• split the event should not repeat pathologically into e2e and core
• Tolerate API downtime in upgrade disruption tests on single-node topologies
• Revert "test/extended/prometheus: Validate alerting rules"
• move olm tests, owners to separate context
• Revert "stop adding individual serviceaccount namespace groups to SCC for e2e tests"
• Bug 1992596: e2e/cli: More basicresources.sh
• fix e2e namespace event typo
• test/extended/prometheus: Validate alerting rules
• suppress NetworkPodsCrashLooping while we work out whether its a bug or tolerable
• Skipping test Hot deploy for openshift perl image for ARM64
• add library-go alerts request client
• Add another failing unidling test for ovn-kubernetes
• e2e: Wait for an individual imagestream import
• images: port image signature workflow test to OCP4/UBI8
• [sig-arch][Late] operators should not create watch channels very often: be more verbose when failing the expected condition
• Update test-cmd after recent removals in oc
• add alerts as event intervals for viewing
• decide not to sort alerts by name
• generated
• Add tolerance value to debug output
• ignore ClusterNotUpgradeable alert on a TechPreviewNoUpgrade cluster
• test/extended/router: Enable edge-terminated gRPC
• RBAC default rules test: allow new configmap in authenticated user
• Minimise test disruption by removing newest workers in scale down
• bug 2011924: Fix inaccurate test error for NodeReady multiple times.
• update canary tests for 4.10
• Fixed RBAC test and moved the new map to the openshift-config-managed namespace
• restrict visibility on openshift-network-features
• test/extended/operators: Guess at Upgradeable exception removal
• test/extended/operators: Allow early KubeletMinorVersionUnsupportedNextUpgrade
• test/extended/operators: New staff-engineer OWNERS
• Bump apirequestcount upper bounds for 4.10
• tigthen the SCC test for fixed levels, relax for unfixed levels
• Ignore upgrade errors on vSphere HW 13 platformsupported
• ignore ClusterNotUpgradeable alert on a TechPreviewNoUpgrade cluster
• Add myself to OWNERS
• refactor writing data artifacts without behavior change
• Add exception for repeated api check events on single-node
• Update watch thresholds for cluster-monitoring-operator
• Set provider name when platform is vSphere
• kubeconfigs: fix capture loop variable in closure
• add search default image registries test to build e2es
• make update
• Disable vSphere-specific storage tests
• write backend disruption json
• Add baremetal operator bootMacAddress update test
• Oauth server tests: Get Oauth issuer url from discovery instead of route
• Skip test 'clone repository using git:// protocol should clone using git:// if no proxy is configured'
• Bug 2019375: tests: skip some tests for proxy
• Set Azure upper bound to 50-percentile through maximum of 60 minutes buckets
• Set AWS upper bound to 50-percentile through maximum of 60 minutes buckets
• Set Metal upper bound to 50-percentile through maximum of 60 minutes buckets
• test: Add admin ack Upgradeable condition gate test
• Add Oleg as approver
• Add more tests for image policy
• Disable topology tests for vSphere in-tree storage plugin
• Add baremetal operator bootMacAddress update test
• Add probe tests
• CLI tests: Softcode the kubernetes.default cluster IP
• Add parallel test for new baremetal resources
• Add env tests
• Added 2 whereabouts e2e tests
• Bug 2021141: Cluster should allow a fast rollout of kube-apiserver is failing on single node
• correct string matching for backend disruption summary gathering
• Finish basicresources.sh migration
• track alerts as persisted metadata for later aggregation
• Bug 2002776: fix intermittent test failure due to missing tools image
• Disable new build test on disconnected
• Set GCP upper bound to 50-percentile through maximum of 60 minutes buckets
• Set vSphere upper bound to 50-percentile through maximum of 60 minutes buckets
• Set OpenStack upper bound to 50-percentile through maximum of 60 minutes buckets
• Maximum from all 60 minute long sequences takes into accout the actual maximum of apirequestscount
• Migrate apiresources.go and add version
• Fix single-node frontends disruption test tolerated disruption %
• Refresh prometheus OWNERS
• Updating openshift-enterprise-tests images to be consistent with ART Reconciling with https://github.com/openshift/ocp-build-data/tree/5b89f5b601508a0bcc0399fd3f34b7aa2e86e90e/images/openshift-enterprise-tests.yml
• oauth e2e: fix the expected http2 error to match golang 1.17+
• oauth/requestheaders: escape semicolon in URI
• Revert "Finish basicresources.sh migration"
• account for longer tests runs in the watch pathological watch test case
• pkg/synthetictests: add toleration for missing versioned etcd-endpoints cm during upgrade
• Skip imageregistry serial test requiring internet access for disconnected environments
• prevent rare alert gathering panic
• Update baremetal tests owners
• olm: enable test on arm64
• Create new-project without updating kubeconfig
• Switch the cluster-role-reapers test to Serial
• Managed cluster should ensure control plane operator: Allow hypershift kapi proxy
• Bug 2015624: Refactor IBM Cloud skips to External
• builds: upgrade php to 7.4
• image_ecosystem: specify available architectures for each S2I image
• image_ecosystem: update imagestreams
• image_ecosystem: add ubi7 imagestreams and x86-only dotnet
• Bump nfs-server version for multi-arch support
• Relocate test-build-*s2i images for multi-arch support
• Bug 2023238: Skipping Django Test until bug is fixed
• Finish basicresources migration with metal fixes
• Specify default namespace when getting API endpoint
• Allow even greater downtime on Azure single-node upgrade
• Allow disruption in image-registry availability during single node upgrades
• Allow operator-lifecycle-manager-packageserver to not define related objects when external controlplane
• Disable app behind service lb with pdb disruption test on single-node
• cli: use ProxyFromEnvironment for requests
• prometheus: adjust telemetry test for remote_write switch
• Local kubeconfig tests: Allow no master nodes when external controlplane topology
• Do not assume presence of metal3 apis for external controlplanes
• Dont test presence of {authentication,cloudcredentials} operator crd when controlplane is external
• Fix 'oc expose can ensure the expose command' for custom service cidrs
• Fix should only include cluster daemonsets that have maxUnavailable for Hypershift
• Only known images used by tests: Fix for Hypershift
• double space preventing proper reporting for disruption
• Skip must-gather tests when running on Hypershift
• Add test to check kubernetes version
• Move shared OWNERS file into generated dir
• test/extended/util/openshift/clusterversionoperator: Delegate admin-ack ownership
• test/extended/util/disruption: Include tolerated percentage when failing
• Skip oc version test in disconnected
• test/e2e/upgrade/adminack: Poll gates for duration of update
• Extutil: Add helpers to get a Hypershift management CLI
• Fix OLM should have imagePullPolicy in Hypershift: Talk to the management cluster
• OLM: Fix tests that exec into pods when using external controlplane
• Fix etcd leader changes are not excessive for Hypershift
• Revert "Add test to check kubernetes version"
• Revert "Skip oc version test in disconnected"
• test/extended/util/openshift/clusterversionoperator: Simplify Fail(f) calls
• test/extended/util/openshift/clusterversionoperator: Return errors
• test/extended/util/openshift/clusterversionoperator: Drop waitForUpgradeable
• vendor: update openshift/kubernetes to include gcp quota fix
• Alter junit test suite names for jobs with multiple junit xmls.
• Bug 2026806: clusterversionoperator/adminack.go: Check for nil cm map
• refactor disruption suite to collapse to a single implementation
• clarify names is disruption output
• refine disruption readability
• Disable intree vsphere tests
• Revert "refactor disruption suite to collapse to a single implementation"
• test/extended/etcd: cleanup OWNERS
• ensure that we always produce output for successful pod sandbox tests
• Fix TestFIPS on Hypershift
• Hypershift: Skip Prometheus should have important platform topology metrics
• force a new payload
• Revert "Alter junit test suite names for jobs with multiple junit xmls."
• test: avoid double slash in path
• Set e2e-images-update label
• Add workaround for baremetal resources test
• Add GCE back into the frontend test
• oc label will properly report removed label
• Revert "Revert "refactor disruption suite to collapse to a single implementation""
• extend disruption timeout to 5 seconds to match old behavior
• fix disruption sampler
• tweak disruption timeouts to get clearer error messages about where we failed
• relax disruption constraints to allow merging more reliable sample
• correct total duration calculation for event intervals
• collapse service LB with PDB into standard disruption check
• update service loadbalancer to explicitly bypass SCCs instead of implicitly by test name to namespace name indirection
• mark the service load balancer test as needing privileged
• extend tcp dial timeout and http response header timeout
• find disruption backends from the event intervals instead of a hardcoded list
• update vendor
• cluster: use newer imagestreams in quickstarts
• image_ecosystem: remove nodejs:12-ubi8
• jenkins: update templates for nodejs:14-ubi8
• Disable etcd test for 1.23
• extend timeouts with 5s for each part
• pkg: Buffer signals for abortCh
• add ingress backends to the stable list
• create late-binding host calculation for disruption
• Bug 2026488: Skip Duplicate openshift-controller-manager Events
• test/extended/util/annotate/generated: Delegate to team leads
• set allowed disruption based on historical p95 data
• disruption: use proxy from environment when measuring disruption
• allow skipping a backend test
• Bug 1981867: add backoff retries for oc explain tests
• Update list of resources for oc explain tests
• update p95 values
• Add sync tests for ErrImagePull monitor intervals.
• Always include event intervals in junit sysout even if no failures.
• add topology to the p95 disruption values
• make missing p95 history obvious
• round the disruption before comparsion so the messages match
• Revert "Add workaround for baremetal resources test"
• stop flaking on disruption because we always produce data for analysis
• Fix vsphere testing workarounds
• switch prometheus queries to use our external route, not an exec pod
• avoid requiring a valid client before running tests
• allow wardle aggregated apiservers to be down pending bug
• Check for presence of metal3 pods on specifc non-baremetal platforms
• add support for build csi volume source
• watch specific alerts for pending and firing
• separate out more alerts into specific tests
• write out zero duration for interesting alerts
• move junit types to eliminate cyclical dep
• Display ErrImagePull messages.
• for upgrades, skip alerts with individual invariant tests
• allow burn budget to flake until kube-apiserver team finds and fixes it
• Remove the workaround for event spam
• Bump k8s deps and o/k
• go mod tidy && go mod vendor
• React to bump
• bump: cadvisor to upstream v0.43.0
• Update image tags and regenerate
• make update
• Update extended test images
• Update test pod bindata to use new allowed images
• Add back missing busybox image
• bump(k8s.io/kubernetes)
• Alter junit test suite names for jobs with multiple junit xmls.
• Bug 1931594: add one more timeout condition for --request-timeout test
• make it easy to create multiple flavors of interval charts
• split watchdog test into its own test
• allow promtail restarts as designed
• Recognize more patterns for image pulls
• Add test for OS updates being staged within 5-10m.
• separate prometheus watch alerts
• baremetal: add test helper
• baremetal: add skip inspection test
• Add support to scale test for baremetal platform
• Allow KA guard probe to fail as designed
• update disruption times and shift to p99
• test/extended: re-enable API data in etcd test
• Sync examples
• image_ecosystem: use unique tags in test logs
• image_ecosystem: sync s2i image tags
• loki moved namespaces, so we have to move to match
• also tolerate guard pod issues for kcm and ks
• etcd: update list of APIs removed
• allow loki target to be down
• adds a synthetic test for checking if the pod.spec.nodeName field is immutable
• storage operator can go upgradeable=unknown if vsphere is unreachable
• baremetal: remove wait condition when deleting extraworkers
• Flake failed sandboxes from bug in new guard pods
• Add X-CSRF-Token header to prevent misleading warning
• Bug 2040793: Fix ordering issues with snapshot e2e
• baremetal: add test to check the presence of preprovisioning images for every deployed worker
• Ensure CAO is stable before running the test
• Adding new cloud provider e2e for alibabacloud
• alert: allow apiserver burn rate to fire
• use real alert data to determine flake and failure timeframes for alerts
• add more frequently failing loki alerts to the allow-list
• Flake instead of fail the timely OSUpdateStaged test on metal and ovirt.
• Revert "add support for build csi volume source"
• operators: extend test for exposed cluster services by optional path
• prometheus: /graph should be inaccessible, test /api/v1/targets
• don't double report early alert failures
• add support for build csi volume source
• Improve edge cases error message for OSUpdateStaged test.
• Debugging hack for missing event intervals.
• Skip some HAProxy tests on FIPS
• Revert "Skip some HAProxy tests on FIPS"
• test/extended/util/url: Use curl --connect-to
• Fix HAProxy tests on FIPS properly
• Updating openshift-enterprise-tests images to be consistent with ART Reconciling with https://github.com/openshift/ocp-build-data/tree/0059e470a61eba28225165168b22d9e77c33c926/images/openshift-enterprise-tests.yml
• Add baremetal high availability tests
• test/extended/router: Dump crypto-policy for FIPS
• Bug 2044824: Update k8s vendoring
• Bug 1825034: IBMCloud: Concurrent CRB create
• Fix broken console disruption test.
• images: update registry to 2.8.0-beta.1
• Only run parallel tests in the canary job since it runs in parallel
• Delete extra workers before running metal high availability test
• update p99 values for disruption and alerts
• tweak the magic numbers for 'no data' to be easier to find
• Revert "Flake failed sandboxes from bug in new guard pods"
• Alibaba: enable init for alibabacloud provider
• add job type key to the error we track for statistical jobs
• Break out test for OSUpdateStaged event with no OSUpdateStarted.
• standardize parsing and selection of historical data
• allow inexact matches for disruption data
• Allow a single test to override suite timeout
• requestheader test must wait for authentication operator to settle after config update
• Revert "Disable intree vsphere tests"
• fail on static pod lifecycle failures
• Reenable wait on worker deletion and increase serial test timeout
• ignore ip-reconciler failures in CI for now
• Improve ignore for ip-reconciler KubeJobFailed alert.
• update alert thresholds after extending static pod install duration
• vsphere is firing this alert/VSphereOpenshiftNodeHealthFail and it is hiding other alerts
• Increase cluster-monitor-operator watch count threshold on single-node clusters
• allow slow kubelet updates for static pods, but still fail if they never succeed
• Set ovnkube-node readiness failure to flake for now.
• Add a separate test for ovn node readiness probe
• Consolidates some constants
• Rearrange test cases and add more post-mortem information
• Fix ovnkube-node readiness test to include successes.
• Increase scale test timeout to 30 minutes
• images/tests: yum update python3-six to workaround ART issue
• Bug 2059299: IBMCloud: Ignore CRB already exists
• Attempt to fix lost events with improved watch resource version handling.
• Make a test to specifically look for image pull backoff for registry.redhat.io
• Switch to batch/v1 CronJob
• MGMT-9440: Fix single node serial tests API crash.
• Update test/extended/apiserver/resiliency.go
• add text to the timelines when you hover
• Use common threshold constant as other event tests
• cleanup network policy ACL extended test
• Add pod securityContext sysctl whitelist tests
• operators should not create watch channels very often: bump OpenStack monitoring operator
• only add failure when no event was found for the target revision
• Bug 2057990: Add debug info for signature test
• build-quota test: learn more about cgroupsv2
• Add DNS availability upgrade test
• Add networking test for invalid external gateway
• Mark image pull from registry.redhat.io as flake to track but not fail
• Give each interval chart a title
• Replace event monitor custom watch/rv logic with Reflector.
• Add early test to check router pods are not on same node
• Pod scheduling tests improve pod deployment identification
• Accept both 401 snd 403 status code for querying Prometheus/Thanos endpoint without authentificaiton
• OWNERS: remove former team member
• OWNERS: monitoring team owns prometheus test helpers
• Clarify that the image mirroring requirement applies to conformance tests only
• Migrate tests away from using samples-operator provided imagestreams
• Migrate builds tests away from using sample imagestreams since samples will be optional
• bump(client-go): need imageregistry config
• add test for imageregistry redirect, this important for performance
• Implementation for egress-router-cni e2e ci test
• Push own blob in image registry redirect test
• Filter out ovn readiness event during installation
• expose errors for repeated event evaluation
• prune jenkins e2e to only validate pipeline strategy builds
• removed unused artifacts
• Attempt multiple different token requests
• add intervals for pod lifecycle
• include pod lifecycle events in intervals
• stop ignoring pods that are created before the test begins
• track every pod in the monitor
• have runonce pods display as gone when their containers are stopped
• order the pod lifecycle by namespace
• make separate pages for pod lifecycle
• tidy up deleted runonce pod end time
• sanitize start/end of pod lifecycle
• bound container lifecycle based on container status termination time
• fix pod missed pod creation start
• include all container waiting events
• Initial implementation for egressfw e2e CI test
• Update where the canary test lists is retrieved from
• create container lifecycle events on observing pod creates
• Revert "Implementation for basic egressfw e2e CI test"
• Flake on FailedCreatePodSandbox if missing ovn readiness file.
• Implementation for basic egressfw e2e CI test
• bump: k8s.io/kubernetes
• coarse serialization of storage tests
• make update
• Check token response error code
• Revert "bump: k8s.io/kubernetes"
• Bump allowed watches for node-tuning-operator
• Increasing BUILD_LOGLEVEL to coincide with adding --quiet to buildah
• test: update synthetic test node restart count
• extended/test/router: omit DNS managed conditions for shards
• Fix typo in OVN Kubernetes skipper
• feat: added version to check to trigger only when we upgrade to major version 4.11
• bump: k8s.io/kubernetes
• make update
• Use ServerGroupResources instead deprecated ServerResources
• increasing build config loglevel
• feat: updated with suggestions
• Add missing operators from the known operators list
• test: bump apirequests upper bounds for VSphere and AWS single node
• Ignore FailedCreatePodSandBox event that is expected to happen
• Revert "Implementation for basic egressfw e2e CI test"
• WIP - showing Jon how to add a test