-
Notifications
You must be signed in to change notification settings - Fork 72
WIP: Sync upstream commits to the 4.11 release branch #267
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
WIP: Sync upstream commits to the 4.11 release branch #267
Conversation
…thub/workflows/release.yaml Signed-off-by: Joe Lanford <[email protected]> Upstream-repository: operator-registry Upstream-commit: a629d38a2273aa0cd43b7f06702c8c631401f30c
Upstream-repository: operator-registry Upstream-commit: 3c2677ea67da9bb96d9f1db9e45dd4e46b81dd8e
This PR bumps operator-framework/api to v0.11.0 Signed-off-by: Anik Bhattacharjee <[email protected]> Upstream-repository: operator-registry Upstream-commit: acaac87c522f1826dce8669eb9108ccacc73af75
In PR openshift#888, there's vendoring of the cel-go package, one of which is named "debug" (github.com/google/cel-go/common/debug). Due to the .gitignore file ignoring all [Dd]bug folders, the vendored package isn't being checked in by git, causing the build to fail. This PR removes the instruction to ignore [Dd]bug/ folders. Signed-off-by: Anik Bhattacharjee <[email protected]> Upstream-repository: operator-registry Upstream-commit: 7a09b48dba29e8d063916f276740a1d737b40121
…openshift#888) * feat(constraint): Add constraint type validation to bundle validation A new type of dependency named constraint (olm.constraint) which uses CEL expression is introduced in OLM. This new type is added to dependencies.json file in the bundle just like gvk and package dependencies. As a result, a new validation for constraint type is added to ensure this type will be validated for bundle dependencies. Signed-off-by: Vu Dinh <[email protected]> Vendor the constraints pkg from api repo Signed-off-by: Vu Dinh <[email protected]> * Add unit test for cel-go constraint validation Signed-off-by: Vu Dinh <[email protected]> Upstream-repository: operator-registry Upstream-commit: 384af6b9235dc718268cb3e148a8c5f2c58e64bc
Updating for reader experience. Signed-off-by: Shivkumar13 <[email protected]> Upstream-repository: operator-registry Upstream-commit: 611525be3bd7374d1e55d43c519e5664e0e91c2c
This commit bumps operator-framework/operator-registry requirement go-sqlite3 from v1.10.0 to v1.14.10 Signed-off-by: Noah Sapse <[email protected]> Co-authored-by: Noah Sapse <[email protected]> Upstream-repository: operator-registry Upstream-commit: 4e3865a5cf1a2b1048d3752071ae1e932953f29c
…ift#897) * feat(registry): add podman auth.json as an alternate location Adds podman default credential location as an alternate location when loading the config for containerd registry resolver Signed-off-by: Jennifer Power <[email protected]> * fix: remove named return values in resolver and add xdg library for cross-platform option Signed-off-by: Jennifer Power <[email protected]> * fix: update go.sum for with new xdg import Signed-off-by: Jennifer Power <[email protected]> Upstream-repository: operator-registry Upstream-commit: c056ac1f075b38031263893758fcfe6c96dcee79
Move the `opm alpha generate` command up to `opm generate` Signed-off-by: kevinrizza <[email protected]> Upstream-repository: operator-registry Upstream-commit: d69d2e41d74c16806fbc7b1ebd516ea6723f04f8
Upstream-repository: operator-registry Upstream-commit: 3fa02bc295358a6e3259a1da11bfa35c89e58898
…Only mode (openshift#905) opm diff in headsOnly mode with an include config will prune all empty channels from a package which can leave the output config with an empty package. This change will remove any empty packages after the channel pruning. Fixes openshift#904 Signed-off-by: Jennifer Power <[email protected]> Upstream-repository: operator-registry Upstream-commit: f7e27734f3a633ec03e55779317dd1af4d2cfd0a
…fy (openshift#898) * fix(registry): deprecated skip-tls and add use-http and skip-tls-verify The skip-tls flag currently uses HTTP instead of the expected behavior of skipping TLS cert validated with HTTPS registries. The new flags seperate untrusted HTTPS from HTTP registries and behave as expected Signed-off-by: Jennifer Power <[email protected]> * chore: add helper function to validate and get TLS options for opm command Signed-off-by: Jennifer Power <[email protected]> * chore: add warning message for plain HTTP in registy add command Signed-off-by: Jennifer Power <[email protected]> Upstream-repository: operator-registry Upstream-commit: 354cd385167844295a3288b5a8e7fb5efb9e652b
…onfigs (openshift#914) Signed-off-by: Joe Lanford <[email protected]> Upstream-repository: operator-registry Upstream-commit: 0b008deff332346247624ccc54df64b75cfe9777
…ontents (openshift#915) Signed-off-by: Joe Lanford <[email protected]> Upstream-repository: operator-registry Upstream-commit: 7827c48d1a09fe320af46afe5e19491784b7f89e
Signed-off-by: Ankita Thomas <[email protected]> Upstream-repository: operator-registry Upstream-commit: d59081db05ae040ebacf74f3ccfefd3bbdd9cdca
Signed-off-by: akihikokuroda <[email protected]> Upstream-repository: operator-registry Upstream-commit: 5566e4b6832a7fc08c12d3c79fc0a0b8c6a2e7aa
* Upgrade github.com/docker/cli to v20.10.12 Signed-off-by: Jan Boehm <[email protected]> * Synchronize docker config loading This prevents data races during testing. The root cause are unsecured globals in the docker cli lib. Signed-off-by: Jan Boehm <[email protected]> * Mention upstream issue that fixes the data race Signed-off-by: Jan Boehm <[email protected]> Upstream-repository: operator-registry Upstream-commit: c496cf028568091b18633a9ee59dadcc6d4b4edb
This commit updates the operator-registry's go version to v1.17. Signed-off-by: Alexander Greene <[email protected]> Upstream-repository: operator-registry Upstream-commit: c970d617ce2d77c096102e2d881a15ef680bacf9
Signed-off-by: akihikokuroda <[email protected]> Upstream-repository: operator-registry Upstream-commit: 5c3f4a3a3365c6684f78a3617e4cfc650bec73e3
…#908 (openshift#909) Signed-off-by: Ryan Emerson <[email protected]> Upstream-repository: operator-registry Upstream-commit: 9999f79682d0073bf9ed5edf2cca1476abf2fc9a
This commit introduces a series of unit tests that ensure that GRPC connections are configurable through the use of the GRPC_PROXY environment variable. Signed-off-by: Alexander Greene <[email protected]> Upstream-repository: operator-lifecycle-manager Upstream-commit: a68dc4262ccb41cc6031ab4ef3823f6ce1713036
…r init hooks (#2523) Signed-off-by: Per G. da Silva <[email protected]> Upstream-repository: operator-lifecycle-manager Upstream-commit: b414e6e881066115b4c5dc63840e8d47c624a405
Prepend the https://github.com to ensure that markdown rendered links are not resolved as local links. Signed-off-by: timflannagan <[email protected]> Upstream-repository: operator-lifecycle-manager Upstream-commit: bfe8f0f5339c03c4f1ac08ca343b0a5ab28a787a
…(#2463) Update any relevant workflows and downgrade the checkouts action from the v2 to v1 version. This change is predicated on the problematic checkouts@v2 behavior that can lead to an increased change of regressing when landing changes to the default branch (e.g. master/main). This is because there's an existing bug in the v2 world where PRs can checkout the wrong HEAD commit, leading to scenarios where you're testing against an out-of-date commit compared to what's in the master branch. A concrete use case is when a PR is opened, and that PR checks out the current HEAD commit from master, and then a change lands in the master branch, and any subsequent workflow retries for that PR will result in that previous HEAD commit being checked out. There are workarounds that exist, like rebasing the existing PR, or like explicitly pointing the ref to the default branch, but this interaction isn't ideal with workflows are triggered on pushes to branches on top of reacting to pull request triggers. The simplest solution is to revert back to the v1 version and find a better longer term solution. Signed-off-by: timflannagan <[email protected]> Upstream-repository: operator-lifecycle-manager Upstream-commit: a2142c1c3b38bd7f0d2a6fb57d6ffdf5e5434604
Signed-off-by: timflannagan <[email protected]> Upstream-repository: operator-lifecycle-manager Upstream-commit: 3f3f00644ca4a50d845db2b97c12afcea56fd5d8
Signed-off-by: Vu Dinh <[email protected]> Upstream-repository: operator-lifecycle-manager Upstream-commit: 4daeb114ccd56cee7132883325da68c80ba70bed
Run the e2e GitHub workflow against master periodically, twice per day, at 00:30 and 12:30 ET. Signed-off-by: Nick Hale <[email protected]> Upstream-repository: operator-lifecycle-manager Upstream-commit: 86a7f54652ecb2fdcc40404cb6afa1660f677ef6
Signed-off-by: Alexander Greene <[email protected]> Upstream-repository: operator-lifecycle-manager Upstream-commit: c6cc1d7f706d355e34b1fc6523693bc13ceb4650
* Introduce golangci-lint support Introduce a root .golangci-lint.yaml file to the repository. This file is responsible for housing the golangci-lint configuration, and the various aggregate linters that are enabled/disabled. Signed-off-by: timflannagan <[email protected]> * .github/workflows: Run the golangci-lint action during the sanity checks Update the sanity workflow and add the golangci-lint action. Signed-off-by: timflannagan <[email protected]> * Fix linting violations outlined by golangci-lint Signed-off-by: timflannagan <[email protected]> Upstream-repository: operator-lifecycle-manager Upstream-commit: b3cded1d82f74feb0879ccc68abc1388c6ca4e70
Signed-off-by: Per G. da Silva <[email protected]> Upstream-repository: operator-lifecycle-manager Upstream-commit: bee0b688521e79a1bba8c2b3463a13a33ae7cd59
/retest-required |
Alright it looks like the github status' were stuck and closing/re-opening worked 🤷 |
/retest-required |
1 similar comment
/retest-required |
/retest |
/test all |
After swapping out the e2e tests from AWS -> GCP we finally got some e2e coverage. The latest e2e run only had a single failure:
Edit: I was able to reproduce this locally and this appears to be the same issue we saw in b638f1c. Going to respin e2e again and investigate this test case failure more locally. /retest-required |
/retest-required |
4 similar comments
/retest-required |
/retest-required |
/retest-required |
/retest-required |
@timflannagan: The following tests failed, say
Full PR test history. Your PR dashboard. Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. I understand the commands that are listed here. |
This PR upgrades the version of yq used in the makefile from v3@latest to [email protected] Upstream-repository: api Upstream-commit: b527a19c8e14e2249d5a5cdb88063cfb136fdafb
This PR upgrades the version of yq used in the makefile from v3@latest to [email protected] Upstream-repository: api Upstream-commit: b527a19c8e14e2249d5a5cdb88063cfb136fdafb
This PR upgrades the version of yq used in the makefile from v3@latest to [email protected] Upstream-repository: api Upstream-commit: b527a19c8e14e2249d5a5cdb88063cfb136fdafb
This PR upgrades the version of yq used in the makefile from v3@latest to [email protected] Upstream-repository: api Upstream-commit: b527a19c8e14e2249d5a5cdb88063cfb136fdafb
This PR upgrades the version of yq used in the makefile from v3@latest to [email protected] Upstream-repository: api Upstream-commit: b527a19c8e14e2249d5a5cdb88063cfb136fdafb
This PR upgrades the version of yq used in the makefile from v3@latest to [email protected] Upstream-repository: api Upstream-commit: b527a19c8e14e2249d5a5cdb88063cfb136fdafb
This PR upgrades the version of yq used in the makefile from v3@latest to [email protected] Upstream-repository: api Upstream-commit: b527a19c8e14e2249d5a5cdb88063cfb136fdafb
This PR upgrades the version of yq used in the makefile from v3@latest to [email protected] Upstream-repository: api Upstream-commit: b527a19c8e14e2249d5a5cdb88063cfb136fdafb
Same commit(s) as #264 but adds a commit to the flaky copied CSV test that filters out terminating namespaces to see if that helps alleviate flakiness while running this test in CI (note: difficult to reproduce this locally even on OCP clusters).