of-dev/v1.7.1

Ddsanataradd new Alert that is fired for each VMI that contains an ephemeral

7560d5f9创建于 1月13日历史提交

文件	最后提交记录	最后更新时间
boilerplate	Add the license linter for pkg/monitoring Signed-off-by: avlitman <alitman@redhat.com>	1 年前
builder	feat: bump code-generator to v0.32.5 and adjust mockgen command Signed-off-by: YuJack <jk82421@gmail.com> (cherry picked from commit 74534beb4fd1adbc99202947377431255ec680f9)	7 个月前
common-instancetypes	Automated update of common-instancetypes bundles to v1.5.1 release-note Updated common-instancetypes bundles to v1.5.1 Signed-off-by: kubevirt-bot <kubevirtbot@redhat.com>	7 个月前
gen-client-python	Remove invalid email reference Replace old bot user email with new kubevirtbot@redhat.com. Also swap git user.name and user.email to use correct defaults. The old email was removed from the kubevirt-bot GitHub account, thus the publish job for nightly build failed: https://prow.ci.kubevirt.io/view/gs/kubevirt-prow/logs/periodic-kubevirt-push-nightly-build-main/1669510740576309248#1:build-log.txt%3A1510 Signed-off-by: Daniel Hiller <dhiller@redhat.com>	2 年前
gen-swagger-doc	apidocs: remove gradle build and properties files Signed-off-by: Brian Carey <bcarey@redhat.com>	1 年前
linter	lint: Add pkg/virt-launcher/virtwrap/access-credentials to linter Added the package pkg/virt-launcher/virtwrap/access-credentials to linter and run "make gofumpt". Signed-off-by: Andrej Krejcir <akrejcir@redhat.com>	8 个月前
prom-metric-linter	Create docs/observability folder with metrics.md Signed-off-by: João Vilaça <machadovilaca@gmail.com>	1 年前
prom-rule-ci	add new Alert that is fired for each VMI that contains an ephemeral hotplug volume. Change label to annotation so we can more easily report the affected volumes. add new prometheus unit test for alert. Signed-off-by: dsanatar <dsanatar@redhat.com>	5 个月前
OWNERS	Remove usage of "filters" in OWNERS The documentation states: "WARNING: The approve plugin does not currently respect filters. Until that is fixed, filters should only be used for the labels key (as shown in the above example)" From now on we can only use OWNERS files per directory. Signed-off-by: Luboslav Pivarc <lpivarc@redhat.com>	1 年前
autobump-generic.sh	Use the term 'main' where possible Signed-off-by: Stu Gott <sgott@redhat.com>	4 年前
autobump-kubevirtci.sh	Use the term 'main' where possible Signed-off-by: Stu Gott <sgott@redhat.com>	4 年前
bazel-build	Change build environment to execute ginkgo in parallel Functional tests on non-kind providers on CI should run the whitelisted tests in paralle. Local tests should still run not parallel. Signed-off-by: rmohr <rmohr@redhat.com>	5 年前
bazel-build-and-fmt.sh	sandbox: Let the bootstrapper configure the new toolchain and ensure that bootstrapping does not take place on pure go targets. Signed-off-by: Roman Mohr <rmohr@redhat.com>	4 年前
bazel-build-functests.sh	passt: Unexport passt binary Not needed anymore, since we wrap it with a container image. This commit unexport it and also remove it from the release. Signed-off-by: Or Shoval <oshoval@redhat.com>	11 个月前
bazel-build-images.sh	Adapt synchronization-controller oci_image name We need it prefixed with "virt-" following https://github.com/kubevirt/kubevirt/pull/15741 but in any case that would've been the standard (see other image target names) Otherwise, this gets pushed to quay.io/kubevirt/synchronization-controller which may be regarded as breaking an API Signed-off-by: Alex Kalenyuk <akalenyu@redhat.com>	8 个月前
bazel-build.sh	passt: Unexport passt binary Not needed anymore, since we wrap it with a container image. This commit unexport it and also remove it from the release. Signed-off-by: Or Shoval <oshoval@redhat.com>	11 个月前
bazel-coverage-report.sh	coverage: create basic coverage report Example: WHAT=//pkg/monitoring/... make coverage-report will create a basic HTML coverage report. Output path to file is printed on stdout. Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前
bazel-fmt.sh	fix typos. Signed-off-by: yojay11717 <lanyujie@inspur.com>	2 年前
bazel-generate.sh	Gazelle: Remove exclude for vendored grpc Includes generated BUILD files Signed-off-by: Luboslav Pivarc <lpivarc@redhat.com>	8 个月前
bazel-goveralls.sh	Move coverage reports over to bazel coverage reports were the last part where we had to compile kubevirt with pure go-code. With this change it is possible to remove "libvirt-devel" from the builder image. Now no build/test dependencies are installedin the build-image which can get out of sync with bazel. Signed-off-by: Roman Mohr <rmohr@redhat.com>	5 年前
bazel-push-images.sh	build: update PUSH_TARGETS to generate list of images for hack/push-container-manifest.sh Signed-off-by: Brian Carey <bcarey@redhat.com>	8 个月前
bazel-race.sh	hack: Introduce bazel-race to generate This commit fixes race detection within the unit tests by explicitly enabling the detector for each target. This replaces the use of the now unsupported --features race command line argument [1]. The replacement command line argument of --io_bazel_rules_go//go/config:race is not used however as it currently complains of cgo not being used when not explicitly enabled by a target [2]. buildozer is used to automate this as part of the generate target. [1] bazel-contrib/rules_go#2671 [2] bazel-contrib/rules_go#3548 Signed-off-by: Lee Yarwood <lyarwood@redhat.com>	8 个月前
bazel-server.sh	add s390x support for kubevirt builder. Signed-off-by: VamsikrishnaSiddu <vamsikrishna.siddu@ibm.com>	2 年前
bazel-test.sh	hack: Introduce bazel-race to generate This commit fixes race detection within the unit tests by explicitly enabling the detector for each target. This replaces the use of the now unsupported --features race command line argument [1]. The replacement command line argument of --io_bazel_rules_go//go/config:race is not used however as it currently complains of cgo not being used when not explicitly enabled by a target [2]. buildozer is used to automate this as part of the generate target. [1] bazel-contrib/rules_go#2671 [2] bazel-contrib/rules_go#3548 Signed-off-by: Lee Yarwood <lyarwood@redhat.com>	8 个月前
bootstrap-ginkgo.sh	Fix ginkgo bootstrapping script errors Some logical errors were present in this script: - Does not check all folders - Path to ginkgo binary wrong Signed-off-by: Alex Kalenyuk <akalenyu@redhat.com>	1 年前
bootstrap.sh	build: Run make generate for the basic TDX enablement Signed-off-by: Zhenchao Liu <zhencliu@redhat.com>	7 个月前
build-copy-artifacts.sh	Fix build on platforms with bash not under /bin/ Not every platform has bash under /bin. For example, nixos keeps all binaries under /nix/store and sets PATH accordingly to point to all of them. In this environment, the right way to access bash is using /usr/bin/env to locate shell. (The only shell guaranteed to exist at a particular path on posix systems is /bin/sh.) /usr/bin/env is always present on all systems so swapping direct call to /bin/sh with /usr/bin/env bash should work everywhere.	7 年前
build-ginkgo.sh	Fix ginkgo bootstrap script We no longer have the ginkgo cli since https://github.com/kubevirt/kubevirt/commit/9d9b35ec9909f126f6836bbf661ae4fc6831fbec so this script silently fails. Signed-off-by: Alex Kalenyuk <akalenyu@redhat.com>	2 年前
build-go.sh	hack/build-go: fix clean Since the transition to go workspace mode in KubeVirt[1] make clean has been broken due to an incompatibility between go clean with -mod=vendor and go.work. This commit unsets -mod=vendor if the target is set to clean. [1] https://github.com/kubevirt/kubevirt/pull/14485 Signed-off-by: Adi Aloni <aaloni@redhat.com>	11 个月前
build-manifests.sh	virt-operator: Remove support for *_SHASUM env variables This support was deprecated by 9376f95a1d5e699e2c440453477aa1573470fd6e some time ago and can be removed now in favor of image tags. This should allow us to better handle heterogeneous clusters as we can now just focus on using a single image manifest backed version tag, offloading architecture specific decisions to other components. The cluster-patch Makefile target and associated logic that used these env variables to patch a local kubevirtci env has also been dropped. Signed-off-by: Lee Yarwood <lyarwood@redhat.com>	11 个月前
build-verify.sh	kubevirtci: replace kci/cluster/virtctl.sh calls KubeVirtCI project has a circular dependency on the virtctl binary [1] - it expects the binary to be present at a certain path. Thus we create our own version of virtctl.sh in the hack folder. We then replace all calls to virtctl.sh with the new location. [1]: https://github.com/kubevirt/kubevirtci/issues/1277 Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前
bump-distroless.sh	Introduce buildifier-prebuilt to reduce rules_go dependency complexity Use buildifier and buildozer as toolchain, to only use download and use the binaries from the host platform (instead of downloading all) and only make it independent of rules_go for simplifying dependency management. Signed-off-by: Roman Mohr <rmohr@google.com>	1 年前
bump-kubevirtci.sh	Adjust bump script to use tagged kubevirtci releases Signed-off-by: Roman Mohr <rmohr@redhat.com>	5 年前
check-for-binaries.sh	testdata/vhostmd0 is more a text file than a binary Signed-off-by: Roman Mohr <rmohr@redhat.com>	5 年前
check-unassigned-tests.sh	Fix check-unassigned-tests script check-unassigned-tests checks that there are no tests that are not assigned to any sig. To do that it ensures that every tests has [sig-<sig_name>] pattern in the test descriptor. This is not enough, since our automation/test.sh script is not using file name based filtering anymore, but label filtering. This means that, in case a test has the [sig-<sig_name>] string in its name, but not the label, the check-unassigned-tests.sh passes but the test is actually not executed. Let's introduce the double check: the legacy file name based check and the new label check. Also, the sig-compute-migrations was not covered by this check, so introduced it too. Signed-off-by: fossedihelm <ffossemo@redhat.com>	1 年前
cluster-build.sh	kubevirtci: move into it's own folder to allow OWNERS We use [kubevirtci] for spinning up virtualized k8s clusters for testing (locally and on CI). Therefore we source the cluster-up folder into kubevirts' project folder root. Also we have the cluster-up-sha.txt which is used to verify which version we use currently, so that we can determine whether an update is necessary. Related to how OWNERS files work, and since both of those items are located in the root folder, we currently need a root approver for approving pull requests that bump kubevirtci. Thus, to be able for sig-buildsystem to own kubevirtci we move it to it's own folder kubevirtci. In that folder we place an OWNERS file with which we as sig-buildsystem claim ownership of kubevirtci. [kubevirtci]: https://github.com/kubevirt/kubevirtci Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前
cluster-clean.sh	kubevirtci: move into it's own folder to allow OWNERS We use [kubevirtci] for spinning up virtualized k8s clusters for testing (locally and on CI). Therefore we source the cluster-up folder into kubevirts' project folder root. Also we have the cluster-up-sha.txt which is used to verify which version we use currently, so that we can determine whether an update is necessary. Related to how OWNERS files work, and since both of those items are located in the root folder, we currently need a root approver for approving pull requests that bump kubevirtci. Thus, to be able for sig-buildsystem to own kubevirtci we move it to it's own folder kubevirtci. In that folder we place an OWNERS file with which we as sig-buildsystem claim ownership of kubevirtci. [kubevirtci]: https://github.com/kubevirt/kubevirtci Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前
cluster-deploy.sh	fix(np): install provided np When KUBEVIRT_DEPLOY_NP env is true, both base np and generated np should be installed. Signed-off-by: fossedihelm <ffossemo@redhat.com>	7 个月前
cluster-services-np.yaml	Deploy np in kubevirtci Expose an env variable KUBEVIRT_DEPLOY_NP, which allow deploying the network policies, allowing testing KV with strict network policies. Signed-off-by: fossedihelm <ffossemo@redhat.com>	7 个月前
cluster-sync.sh	Honor external imagePullPolicy setting With https://github.com/kubevirt/kubevirt/pull/15220 the imagePullPolicy is always set to Always, even if a developer explicitly set it to something else. This is wrong, since CI or other might want to use a different policy. Signed-off-by: fossedihelm <ffossemo@redhat.com>	10 个月前
cluster-up.sh	kubevirtci: move into it's own folder to allow OWNERS We use [kubevirtci] for spinning up virtualized k8s clusters for testing (locally and on CI). Therefore we source the cluster-up folder into kubevirts' project folder root. Also we have the cluster-up-sha.txt which is used to verify which version we use currently, so that we can determine whether an update is necessary. Related to how OWNERS files work, and since both of those items are located in the root folder, we currently need a root approver for approving pull requests that bump kubevirtci. Thus, to be able for sig-buildsystem to own kubevirtci we move it to it's own folder kubevirtci. In that folder we place an OWNERS file with which we as sig-buildsystem claim ownership of kubevirtci. [kubevirtci]: https://github.com/kubevirt/kubevirtci Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前
common.sh	kubevirtci: move into it's own folder to allow OWNERS We use [kubevirtci] for spinning up virtualized k8s clusters for testing (locally and on CI). Therefore we source the cluster-up folder into kubevirts' project folder root. Also we have the cluster-up-sha.txt which is used to verify which version we use currently, so that we can determine whether an update is necessary. Related to how OWNERS files work, and since both of those items are located in the root folder, we currently need a root approver for approving pull requests that bump kubevirtci. Thus, to be able for sig-buildsystem to own kubevirtci we move it to it's own folder kubevirtci. In that folder we place an OWNERS file with which we as sig-buildsystem claim ownership of kubevirtci. [kubevirtci]: https://github.com/kubevirt/kubevirtci Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前
config-default.sh	Bump kubevirtci [9fd40b1 owners: Remove davidvossel from approvers as no longer involved](https://github.com/kubevirt/kubevirtci/pull/1562) [543ff5b flannel: Increase max pods per node](https://github.com/kubevirt/kubevirtci/pull/1557) [672dc39 Automatic bump of CentOS Stream to latest](https://github.com/kubevirt/kubevirtci/pull/1558) release-note NONE Signed-off-by: kubevirt-bot <kubevirtbot@redhat.com>	7 个月前
config-kubevirtci.sh	kubevirtci: move into it's own folder to allow OWNERS We use [kubevirtci] for spinning up virtualized k8s clusters for testing (locally and on CI). Therefore we source the cluster-up folder into kubevirts' project folder root. Also we have the cluster-up-sha.txt which is used to verify which version we use currently, so that we can determine whether an update is necessary. Related to how OWNERS files work, and since both of those items are located in the root folder, we currently need a root approver for approving pull requests that bump kubevirtci. Thus, to be able for sig-buildsystem to own kubevirtci we move it to it's own folder kubevirtci. In that folder we place an OWNERS file with which we as sig-buildsystem claim ownership of kubevirtci. [kubevirtci]: https://github.com/kubevirt/kubevirtci Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前
config.sh	kubevirtci: move into it's own folder to allow OWNERS We use [kubevirtci] for spinning up virtualized k8s clusters for testing (locally and on CI). Therefore we source the cluster-up folder into kubevirts' project folder root. Also we have the cluster-up-sha.txt which is used to verify which version we use currently, so that we can determine whether an update is necessary. Related to how OWNERS files work, and since both of those items are located in the root folder, we currently need a root approver for approving pull requests that bump kubevirtci. Thus, to be able for sig-buildsystem to own kubevirtci we move it to it's own folder kubevirtci. In that folder we place an OWNERS file with which we as sig-buildsystem claim ownership of kubevirtci. [kubevirtci]: https://github.com/kubevirt/kubevirtci Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前
conformance.sh	conformance: Fix E2E_LABEL generation Originally this would fail with RUN_ON_ARM64_INFRA enabled with: 09:01:10: + sonobuoy run \ --wait \ --plugin _out/manifests/release/conformance.yaml \ --plugin-env kubevirt-conformance.CONTAINER_PREFIX=registry:5000/kubevirt \ --plugin-env kubevirt-conformance.KUBEVIRT_PROVIDER=kind-1.31 \ --plugin-env 'kubevirt-conformance.E2E_LABEL=(conformance)&&(!(RequiresOutsideConnectivity' '&&' 'IPv6))' 09:01:10: Error: accepts 0 arg(s), received 2 As we always set label_filter the original add_to_label_filter func can be removed and replaced with simple string concatenation. The same style of string concatenation is now also used for sonobuoy_args. The !requires-amd64 label is also added to the filter for arm64. E2E_LABEL is now always passed given we always provide at least "(conformance)" to sonobuoy. Signed-off-by: Lee Yarwood <lyarwood@redhat.com>	10 个月前
coverage-report.sh	coverage: create basic coverage report Example: WHAT=//pkg/monitoring/... make coverage-report will create a basic HTML coverage report. Output path to file is printed on stdout. Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前
coverage.sh	coverage: create basic coverage report Example: WHAT=//pkg/monitoring/... make coverage-report will create a basic HTML coverage report. Output path to file is printed on stdout. Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前
dep-update.sh	chore(deps): Enable go workspace sync Enable synchronization of modules in go workspace by running go work sync after updating dependencies. Run go work sync initially to bring dependencies into a consistent state. Signed-off-by: Felix Matouschek <fmatouschek@redhat.com>	1 年前
deploy-to-nodes.sh	passt: Introduce passt cni daemonSet Use it to deploy the cni binary to the nodes. Signed-off-by: Or Shoval <oshoval@redhat.com>	11 个月前
dockerized	chore: update builder version Signed-off-by: YuJack <jk82421@gmail.com> (cherry picked from commit 37749f222a8c3e3be944e0ca2b286f6a089efb7d)	7 个月前
dump.sh	Use dump logic to diagnose cluster-up issues If CI lanes fail to properly install kubevirt, collect logs to simplify developer flows. Signed-off-by: Roman Mohr <rmohr@redhat.com>	6 年前
fossa.sh	Add branch inside ci run Since inside prow we are in a detached head state we add the branch name. Signed-off-by: Daniel Hiller <dhiller@redhat.com>	3 年前
func-tests-image.sh	Introduce fail_if_cri_bin_missing() in common.sh This way we can fail only in the scripts that depand on KUBEVIRT_CRI variable. Signed-off-by: Itamar Holder <iholder@redhat.com>	4 年前
functests.sh	tests: Fix focus and skip support In case there was whitespace in the focus / skip vars, it was corrupting the command. Using bash array is safer when it involves appending variables, especially when they might have spaces. Usage example: export KUBEVIRT_E2E_FOCUS="should be able to reach the outside world" Note that some characters need to be escaped still: export KUBEVIRT_E2E_FOCUS="should be able to reach the outside world \[IPv4\]" Signed-off-by: Or Shoval <oshoval@redhat.com>	10 个月前
fuzz.sh	Add basic admit webhook fuzzing for VMs and VMIs Signed-off-by: Roman Mohr <rmohr@google.com>	2 年前
gen-proto.sh	Introducing vmi status synchronization controller The synchronization controller synchronizes the newly added migration source and target states between a source and target controller. The source and target controller can be the same controller. Communication happens using the gRPC introduced in this commit. There are 2 functions: - SyncTargetMigrationStatus - SyncSourceMigrationStatus Both take a VMIStatusRequest which has 2 fields: 1. MigrationID, a unique identifier 2. MigrationState This is a copy of the migration state from the sending party. Depending on if the receiver is interested in the source or target only the source or target state is updated by the receiver. An update to a VMI triggers the controller main reconcile loop, which then can trigger either a SyncSource or SyncTarget (or both) call. Added all the needed infra structure to kubevirt operator to deploy the new controller in a separate pod which has its own service account. This service account is part of the list of kubevirt service accounts which is allowed to modify VMI status fields. Added unit tests for synchronization controller demonstrating communication between the same controller (cross namespace) and different controllers (cross cluster). Is part of VEP 24 Decentralized live migration https://github.com/kubevirt/enhancements/issues/24 Signed-off-by: Alexander Wels <awels@redhat.com>	1 年前
generate.sh	VMPool: Introduce v1beta1 version Signed-off-by: svarnam <svarnam@nvidia.com>	7 个月前
go-build-functests.sh	Add s390x to build scripts Signed-off-by: Jan Schintag <jan.schintag@de.ibm.com>	1 年前
gofumpt.sh	Linter: Move covered directories list This is done in order to set up an OWNERS file in the following commit. Signed-off-by: Orel Misan <omisan@redhat.com>	1 年前
golangci-lint.sh	linter: Update the golangci-lint tool command line Between v1 and v2 some arguments have been replaced. Signed-off-by: Edward Haas <edwardh@redhat.com>	10 个月前
gosec.sh	gosec.sh: make it clearer where the output goes It was not trivial for me to find. Signed-off-by: Dan Kenigsberg <danken@redhat.com>	5 年前
goveralls.sh	Simplify ignore files expression This was missing a couple generated files, like zz_generated.defaults.go, deepcopy_generated.go and api.pb.go Signed-off-by: Daniel Hiller <dhiller@redhat.com>	4 年前
integration-test.sh	make: Introduce a new target to run integration tests Functionality which requires OS/kernel services can now be validated in a dedicated target: integ-test. The tests are running in a container (isolated) which is privileged. Tests are expected to be skipped if not explicitly directed to run when executed in this context. i.e. each integration test should include instructions to be skipped if the run-integration-tests argument flag is not set. To run the tests, execute make integ-test. Signed-off-by: Edward Haas <edwardh@redhat.com>	2 年前
kwok-perftests.sh	tests: drop flags.KubeVirtOcPath As much as we love OpenShift, we don't need to test or even use its command line tool oc in KubeVirt. Let us simplify our command line dealings by dropping it. Signed-off-by: Dan Kenigsberg <danken@redhat.com>	1 年前
license-header-check.sh	Add license header linter for virt-handler under pkg/ Signed-off-by: avlitman <alitman@redhat.com>	9 个月前
lint-test-cleanup-label.sh	make, lint: Limit use of OncePerOrderedCleanup The OncePerOrderedCleanup label was introduced in https://github.com/kubevirt/kubevirt/pull/13786 as a pilot for SIG-Network and virtctl. Its purpose is to enable sharing of VMs between multiple tests thus reducing test duration and resource usage. Since the feature is temporarily only allowed in the above areas, this commit adds a linter that fails the lint process should a file located elsewhere use the label. If additional files/areas would like to participate in the pilot, please add required file or directory to the EXCLUDE_PATTERN variable. This entire commit is expected to be reverted once the whole project is on board. Signed-off-by: Nir Dothan <ndothan@redhat.com>	1 年前
manifests.sh	Allow testing with any number of concurrent ginkgo processes This is the only tech debt that remains in order for anyone to be able to run kubevirt tests with any amount of concurrent ginkgo processes (export KUBEVIRT_E2E_PARALLEL_NODES=<NUM> on the host executing the tests). Signed-off-by: Alex Kalenyuk <akalenyu@redhat.com>	1 年前
multi-arch.sh	[multi-arch] Use formatted archname in push-container-manifest bazel-push-images.sh puts the digests files in paths based on the formatted arch names. The same does not happen in push-container-manifest.sh which tries to find digests file using the plain arch names. This results in a failure if you are building aarch64 and x86_64 together, as it happens for example in automation/release.sh Move the format_archname function to common.sh file, so that it can be also used by push-container-manifest.sh Signed-off-by: fossedihelm <ffossemo@redhat.com>	2 年前
olm.sh	Added missing fields in CSV, improved verification Signed-off-by: Marc Sluiter <msluiter@redhat.com>	7 年前
perfscale-tests.sh	Move performance job artifacts to the base artifacts directory Signed-off-by: Ryan Hallisey <rhallisey@nvidia.com>	10 个月前
perftests.sh	Move performance job artifacts to the base artifacts directory Signed-off-by: Ryan Hallisey <rhallisey@nvidia.com>	10 个月前
print-workspace-status.sh	Merge pull request #6199 from ormergi/conformance-pick-container-tag Support passing container tag/digest on Kubevirt Conformance plugin	4 年前
publish-staging.sh	Remove invalid email reference Replace old bot user email with new kubevirtbot@redhat.com. Also swap git user.name and user.email to use correct defaults. The old email was removed from the kubevirt-bot GitHub account, thus the publish job for nightly build failed: https://prow.ci.kubevirt.io/view/gs/kubevirt-prow/logs/periodic-kubevirt-push-nightly-build-main/1669510740576309248#1:build-log.txt%3A1510 Signed-off-by: Daniel Hiller <dhiller@redhat.com>	2 年前
push-container-manifest.sh	build: update PUSH_TARGETS to generate list of images for hack/push-container-manifest.sh Signed-off-by: Brian Carey <bcarey@redhat.com>	8 个月前
realtime-perftests.sh	tests: drop flags.KubeVirtOcPath As much as we love OpenShift, we don't need to test or even use its command line tool oc in KubeVirt. Let us simplify our command line dealings by dropping it. Signed-off-by: Dan Kenigsberg <danken@redhat.com>	1 年前
release.sh	release.sh: relabel the contents of the volumes with a private SELinux context Signed-off-by: Vladik Romanovsky <vromanso@redhat.com>	1 年前
rpm-deps.sh	hack, rpm-deps: Separate rpmtree for passt In order to reduce attack surface in virt-handler by removing unnecessary binaries, separate the passt rpmtree from handler_base. Remove the passt package from handlerbase and define a new rpmtree dedicated to passt, such that the tree is decoupled and can be manipulated separately in following commits. Signed-off-by: Nir Dothan <ndothan@redhat.com>	7 个月前
sync-kubevirtci-stable-provider.sh	fix(check-flakes): avoid curl 429 by reading locally In previous version of repeated_test.sh we did a curl against the kubevirtci version to fetch the latest stable kubevirtci provider, which can cause 429 errors if the curl is executed very often. Thus we move retrieval of the stable provider for a kubevirtci provider into the sync-kubevirtci and write it into the kubevirtci folder into a text file. This we then just retrieve when executing the repeated_test.sh. Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前
sync-kubevirtci.sh	Allow overriding kubevirtci in local env The env var was being hardcoded and thus it was not possible for one to specify a custom tag Signed-off-by: Alex Kalenyuk <akalenyu@redhat.com>	7 个月前
timestamps.sh	Makefile: Add timestamps to make targets Add timestamps to the console output in order to detect where operations time was spent, helping with debugging timeout or slow run issues. The make targets have been wrapped with a customized shell that adds the time stamp on each outputted line (stdout and stderr). The customized shell includes a pipefail option ensuring that the exit status of the pipe that passes stdout/stderr to the timestamps function is propagated instead of being hidden behind the timestamps function that can never fail. Signed-off-by: Daniel Belenky <dbelenky@redhat.com>	5 年前
update-generated-api-testdata.sh	build: run update-generated-api-testdata.sh with -mod=readonly make generate fails with the following: go: ignoring package kubevirt.io/api/apitesting which exists in the vendor directory but is missing from vendor/modules.txt. To sync the vendor directory run go mod vendor. cannot find module providing package kubevirt.io/api/apitesting: import lookup disabled by -mod=vendor (Go version in go.mod is at least 1.14 and vendor directory exists.) Setting -mod=readonly allows the update-generated-api-testdata.sh script to complete successfully. Signed-off-by: Brian Carey <bcarey@redhat.com>	1 年前
verify-generate.sh	generate, Show git diff in case generate diff exists In case generate diff exists, show both the files and the actual diff, it makes it easier and faster to fix in case its a small change. Signed-off-by: Or Shoval <oshoval@redhat.com>	4 年前
verify-rpm-deps.sh	rpm: Move repo.yaml It doesn't make sense to have this in the top-level directory. Signed-off-by: Andrea Bolognani <abologna@redhat.com>	4 年前
version.sh	Add container-tag to bazel builds common values Signed-off-by: Or Mergi <ormergi@redhat.com>	4 年前
virtctl.sh	kubevirtci: replace kci/cluster/virtctl.sh calls KubeVirtCI project has a circular dependency on the virtctl binary [1] - it expects the binary to be present at a certain path. Thus we create our own version of virtctl.sh in the hack folder. We then replace all calls to virtctl.sh with the new location. [1]: https://github.com/kubevirt/kubevirtci/issues/1277 Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前
whatchanged.sh	kubevirtci: move into it's own folder to allow OWNERS We use [kubevirtci] for spinning up virtualized k8s clusters for testing (locally and on CI). Therefore we source the cluster-up folder into kubevirts' project folder root. Also we have the cluster-up-sha.txt which is used to verify which version we use currently, so that we can determine whether an update is necessary. Related to how OWNERS files work, and since both of those items are located in the root folder, we currently need a root approver for approving pull requests that bump kubevirtci. Thus, to be able for sig-buildsystem to own kubevirtci we move it to it's own folder kubevirtci. In that folder we place an OWNERS file with which we as sig-buildsystem claim ownership of kubevirtci. [kubevirtci]: https://github.com/kubevirt/kubevirtci Signed-off-by: Daniel Hiller <dhiller@redhat.com>	1 年前