`PercentageRunnersBusy`, in combination with a secondary `TotalInProgressAndQueuedWorkflowRuns` metric, enables scale-from-zero for PercentageRunnersBusy.
Please see the new `Autoscaling to/from 0` section in the updated documentation about how it works.
Resolves#522
This adds the initial version of ScheduledOverrides to HorizontalRunnerAutoscaler.
`MinReplicas` overriding should just work.
When there are two or more ScheduledOverrides, the earliest one that matched is activated. Each ScheduledOverride can be recurring or one-time. If you have two or more ScheduledOverrides, only one of them should be one-time. And the one-time override should be the earliest item in the list to make sense.
Tests will be added in another commit. Logging improvements and additional observability in HRA.Status will also be added in yet another commits.
Ref #484
Adds two types `RecurrenceRule` and `Period` and one function `MatchSchedule` as the foundation for building the upcoming ScheduledOverrides feature.
Ref #484
Previously any non-go changes resulted in `make docker-build` rerunning time-consufming `go build`. This fixes that by adding clearly unnecessary files .dockerignore
- You can now use `make acceptance/run` to run only a specific acceptance test case
- Add note about Ubuntu 20.04 users / snap-provided docker
- Add instruction to run Ginkgo tests
- Extract acceptance/load from acceptance/kind
- Make `acceptance/pull` not depend on `docker-build`, so that you can do `make docker-build acceptance/load` for faster image reload
This is an attempt to support scaling from/to zero.
The basic idea is that we create a one-off "registration-only" runner pod on RunnerReplicaSet being scaled to zero, so that there is one "offline" runner, which enables GitHub Actions to queue jobs instead of discarding those.
GitHub Actions seems to immediately throw away the new job when there are no runners at all. Generally, having runners of any status, `busy`, `idle`, or `offline` would prevent GitHub actions from failing jobs. But retaining `busy` or `idle` runners means that we need to keep runner pods running, which conflicts with our desired to scale to/from zero, hence we retain `offline` runners.
In this change, I enhanced the runnerreplicaset controller to create a registration-only runner on very beginning of its reconciliation logic, only when a runnerreplicaset is scaled to zero. The runner controller creates the registration-only runner pod, waits for it to become "offline", and then removes the runner pod. The runner on GitHub stays `offline`, until the runner resource on K8s is deleted. As we remove the registration-only runner pod as soon as it registers, this doesn't block cluster-autoscaler.
Related to #447
This makes logging more concise by changing logger names to something like `controllers.Runner` to `actions-runner-controller.runner` after the standard `controller-rutime.controller` and reducing redundant logs by removing unnecessary requeues. I have also tweaked log messages so that their style is more consistent, which will also help readability. Also, runnerreplicaset-controller lacked useful logs so I have enhanced it.
As part of #282, I have introduced some caching mechanism to avoid excessive GitHub API calls due to the autoscaling calculation involving GitHub API calls is executed on each Webhook event.
Apparently, it was saving the wrong value in the cache- The value was one after applying `HRA.Spec.{Max,Min}Replicas` so manual changes to {Max,Min}Replicas doesn't affect RunnerDeployment.Spec.Replicas until the cache expires. This isn't what I had wanted.
This patch fixes that, by changing the value being cached to one before applying {Min,Max}Replicas.
Additionally, I've also updated logging so that you observe which number was fetched from cache, and what number was suggested by either TotalNumberOfQueuedAndInProgressWorkflowRuns or PercentageRunnersBusy, and what was the final number used as the desired-replicas(after applying {Min,Max}Replicas).
Follow-up for #282
Since #392, the runner controller could have taken unexpectedly long time until it finally notices that the runner has been registered to GitHub. This patch fixes the issue, so that the controller will notice the successful registration in approximately 1 minute(hard-coded).
More concretely, let's say you had configured a long sync-period of like 10m, the runner controller could have taken approx 10m to notice the successful registration. The original expectation was 1m, because it was intended to recheck every 1m as implemented in #392. It wasn't working as such due to my misunderstanding in how requeueing work.
We occasionally encountered those errors while the underlying RunnerReplicaSet is being recreated/replaced on RunnerDeployment.Spec.Template update. It turned out to be due to that the RunnerDeployment controller was waiting for the runner pod becomes `Running`, intead of the new replacement runner to have registered to GitHub. This fixes that, by trying to Runner.Status.Phase to `Running` only after the runner in the runner pod appears to be registered.
A side-effect of this change is that runner controller would call more "ListRunners" GitHub Actions API. I've reviewed and improved the runner controller code and Runner CRD to make make the number of calls minimum. In most cases, ListRunners should be called only twice for each runner creation.
This allows you to trigger autoscaling depending on check_run names(i.e. actions job names). If you are willing to differentiate scale amount only for a specific job, or want to scale only on a specific job, try this.
PercentageRunnerBusy seems to have regressed since #355 due to that RunnerDeployment.Spec.Selector is empty by default and the HRA controller was using that empty selector to query runners, which somehow returned 0 runners. This fixes that by using the newly added automatic `runner-deployment-name` label for the default runner label and the selector, which avoids querying with empty selector.
Ref https://github.com/summerwind/actions-runner-controller/issues/377#issuecomment-795200205
We sometimes see that integration test fails due to runner replicas not meeting the expected number in a timely manner. After investigating a bit, this turned out to be due to that HRA updates on webhook-based autoscaler and HRA controller are conflicting. This changes the controllers to use Patch instead of Update to make conflicts less likely to happen.
I have also updated the hra controller to use Patch when updating RunnerDeployment, too.
Overall, these changes should make the webhook-based autoscaling more reliable due to less conflicts.
Similar to #348 for #346, but for HRA.Spec.CapacityReservations usually modified by the webhook-based autoscaler controller.
This patch tries to fix that by improving the webhook-based autoscaler controller to omit expired reservations on updating HRA spec.
`kubectl get horizontalrunnerautoscalers.actions.summerwind.dev` shows HRA.status.desiredReplicas as the DESIRED count. However the value had been not taking capacityReservations into account, which resulted in showing incorrect count when you used webhook-based autoscaler, or capacityReservations API directly. This fixes that.
The controller had been writing confusing messages like the below on missing scale target:
```
Found too many scale targets: It must be exactly one to avoid ambiguity. Either set WatchNamespace for the webhook-based autoscaler to let it only find HRAs in the namespace, or update Repository or Organization fields in your RunnerDeployment resources to fix the ambiguity.{"scaleTargets": ""}
```
This fixes that, while improving many kinds of messages written while reconcilation, so that the error message is more actionable.
We occasionally see logs like the below:
```
2021-02-24T02:48:26.769ZERRORFailed to update runner status{"runnerreplicaset": "testns-244ol/example-runnerdeploy-j5wzf", "error": "Operation cannot be fulfilled on runnerreplicasets.actions.summerwind.dev \"example-runnerdeploy-j5wzf\": the object has been modified; please apply your changes to the latest version and try again"}
github.com/go-logr/zapr.(*zapLogger).Error
/home/runner/go/pkg/mod/github.com/go-logr/[email protected]/zapr.go:128
github.com/summerwind/actions-runner-controller/controllers.(*RunnerReplicaSetReconciler).Reconcile
/home/runner/work/actions-runner-controller/actions-runner-controller/controllers/runnerreplicaset_controller.go:207
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).reconcileHandler
/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:256
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).processNextWorkItem
/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:232
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).worker
/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:211
k8s.io/apimachinery/pkg/util/wait.JitterUntil.func1
/home/runner/go/pkg/mod/k8s.io/[email protected]/pkg/util/wait/wait.go:152
k8s.io/apimachinery/pkg/util/wait.JitterUntil
/home/runner/go/pkg/mod/k8s.io/[email protected]/pkg/util/wait/wait.go:153
k8s.io/apimachinery/pkg/util/wait.Until
/home/runner/go/pkg/mod/k8s.io/[email protected]/pkg/util/wait/wait.go:88
2021-02-24T02:48:26.769ZERRORcontroller-runtime.controllerReconciler error{"controller": "testns-244olrunnerreplicaset", "request": "testns-244ol/example-runnerdeploy-j5wzf", "error": "Operation cannot be fulfilled on runnerreplicasets.actions.summerwind.dev \"example-runnerdeploy-j5wzf\": the object has been modified; please apply your changes to the latest version and try again"}
github.com/go-logr/zapr.(*zapLogger).Error
/home/runner/go/pkg/mod/github.com/go-logr/[email protected]/zapr.go:128
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).reconcileHandler
/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:258
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).processNextWorkItem
/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:232
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).worker
/home/runner/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:211
k8s.io/apimachinery/pkg/util/wait.JitterUntil.func1
/home/runner/go/pkg/mod/k8s.io/[email protected]/pkg/util/wait/wait.go:152
k8s.io/apimachinery/pkg/util/wait.JitterUntil
/home/runner/go/pkg/mod/k8s.io/[email protected]/pkg/util/wait/wait.go:153
k8s.io/apimachinery/pkg/util/wait.Until
/home/runner/go/pkg/mod/k8s.io/[email protected]/pkg/util/wait/wait.go:88
```
which can be compacted into one-liner, without the useless stack trace, without double-logging the same error from the logger and the controller.
Apparently I have mistakenly removed `push` option from the workflow in #323 which resulted in new runner build #323 not being pushed. This fixes that.
Changes:
1. Fix length of github-webhook-server port name
2. Add a cluster role binding for github-webhook-server
3. Remove --enable-leader-election from github-webhook-server
I have heard from some user that they have hundred thousands of `status=completed` workflow runs in their repository which effectively blocked TotalNumberOfQueuedAndInProgressWorkflowRuns from working because of GitHub API rate limit due to excessive paginated requests.
This fixes that by separating list-workflow-runs calls to two - one for `queued` and one for `in_progress`, which can make the minimum API call from 1 to 2, but allows it to work regardless of number of remaining `completed` workflow runs.
This enhances the controller to recreate the runner pod if the corresponding runner has failed to register itself to GitHub within 10 minutes(currently hard-coded).
It should alleviate #288 in case the root cause is some kind of transient failures(network unreliability, GitHub down, temporarly compute resource shortage, etc).
Formerly you had to manually detect and delete such pods or even force-delete corresponding runners to unblock the controller.
Since this enhancement, the controller does the pod deletion automatically after 10 minutes after pod creation, which result in the controller create another pod that might work.
Ref #288
When we used `QueuedAndInProgressWorkflowRuns`-based autoscaling, it only fetched and considered only the first 30 workflow runs at the reconcilation time. This may have resulted in unreliable scaling behaviour, like scale-in/out not happening when it was expected.
* feat: HorizontalRunnerAutoscaler Webhook server
This introduces a Webhook server that responds GitHub `check_run`, `pull_request`, and `push` events by scaling up matched HorizontalRunnerAutoscaler by 1 replica. This allows you to immediately add "resource slack" for future GitHub Actions job runs, without waiting next sync period to add insufficient runners.
This feature is highly inspired by https://github.com/philips-labs/terraform-aws-github-runner. terraform-aws-github-runner can manage one set of runners per deployment, where actions-runner-controller with this feature can manage as many sets of runners as you declare with HorizontalRunnerAutoscaler and RunnerDeployment pairs.
On each GitHub event received, the webhook server queries repository-wide and organizational runners from the cluster and searches for the single target to scale up. The webhook server tries to match HorizontalRunnerAutoscaler.Spec.ScaleUpTriggers[].GitHubEvent.[CheckRun|Push|PullRequest] against the event and if it finds only one HRA, it is the scale target. If none or two or more targets are found for repository-wide runners, it does the same on organizational runners.
Changes:
* Fix integration test
* Update manifests
* chart: Add support for github webhook server
* dockerfile: Include github-webhook-server binary
* Do not import unversioned go-github
* Update README
One of the pod recreation conditions has been modified to use hash of runner spec, so that the controller does not keep restarting pods mutated by admission webhooks. This naturally allows us, for example, to use IRSA for EKS that requires its admission webhook to mutate the runner pod to have additional, IRSA-related volumes, volume mounts and env.
Resolves#200
It turned out previous versions of runner images were unable to run actions that require `AGENT_TOOLSDIRECTORY` or `libyaml` to exist in the runner environment. One of notable examples of such actions is [`ruby/setup-ruby`](https://github.com/ruby/setup-ruby).
This change adds the support for those actions, by setting up AGENT_TOOLSDIRECTORY and installing libyaml-dev within runner images.
docker:dind container creates `/var/run/docker.sock` with root user and root group.
so, docker command in runner container needs root privileges to use docker.sock and docker action fails because lack of permission.
Use tcp connection between runner and docker container, so runner container doesn't need root privileges to run docker, and can run docker action.
Fixes#174
Acceptance tests are passing with the chart. In addition to standard chart values, syncPeriod is supported.
Please use it as a foundation for further collaboration.
Ref #184
Inspired by #91
Related #61
Since the initial implementation of RunnerDeployment and until this change, any update to a runner deployment has been leaving old runner replicasets until the next resync interval. This fixes that, by continusouly retrying the reconcilation 10 seconds later to see if there are any old runner replicasets that can be removed.
In addition to that, the cleanup of old runner replicasets has been improved to be deferred until all the runners of the newest replica set to be available. This gives you hopefully zero or at less downtime updates of runner deployments.
Fixes#24
I had observed athe exact issue seen for the 4th option described in https://github.com/elastic/cloud-on-k8s/issues/1822, which resulted in actions-runner-controller is unable to create nor update runners. This fixes that.
I've also updated README to introduce RunnerDeployment and manually tested it to work after the fix.
---
`actions-runner-controller` has been failing while creating and updating runners:
```
2020-03-05T11:05:16.610+0900 ERROR controllers.Runner Failed to update runner {"runner": "default/example-runner", "error": "Runner.actions.summerwind.dev \"example-runner\" is invalid: []: Invalid value: map[string]interface {}{\"apiVersion\":\"actions.summerwind.dev/v1alpha1\", \"kind\":\"Runner\", \"metadata\":map[string]interface {}{\"creationTimestamp\":\"2020-03-05T02:05:16Z\", \"finalizers\":[]interface {}{\"runner.actions.summerwind.dev\"}, \"generation\":2, \"name\":\"example-runner\", \"namespace\":\"default\", \"resourceVersion\":\"911496\", \"selfLink\":\"/apis/actions.summerwind.dev/v1alpha1/namespaces/default/runners/example-runner\", \"uid\":\"48b62d07-ff2c-42d6-878c-d3f951202209\"}, \"spec\":map[string]interface {}{\"env\":interface {}(nil), \"image\":\"\", \"repository\":\"mumoshu/actions-runner-controller-ci\"}}: validation failure list:\nspec.env in body must be of type array: \"null\""}
github.com/go-logr/zapr.(*zapLogger).Error
/Users/c-ykuoka/go/pkg/mod/github.com/go-logr/[email protected]/zapr.go:128
github.com/summerwind/actions-runner-controller/controllers.(*RunnerReconciler).Reconcile
/Users/c-ykuoka/p/actions-runner-controller/controllers/runner_controller.go:88
```
This seems like the exact issue seen in the 4th option in https://github.com/elastic/cloud-on-k8s/issues/1822
I also observed the same issue is failing while the runnerset controller is trying to create/update runners:
```
Also while creating runner in the runnerset controller:
2020-03-05T11:15:01.223+0900 ERROR controller-runtime.controller Reconciler error {"controller": "runnerset", "request": "default/example-runnerset", "error": "Runner.actions.summerwind.dev \"example-runnersetgp56m\" is invalid: []: Invalid value: map[string]interface {}{\"apiVersion\":\"actions.summerwind.dev/v1alpha1\", \"kind\":\"Runner\", \"metadata\":map[string]interface {}{\"creationTimestamp\":\"2020-03-05T02:15:01Z\", \"generateName\":\"example-runnerset\", \"generation\":1, \"name\":\"example-runnersetgp56m\", \"namespace\":\"default\", \"ownerReferences\":[]interface {}{map[string]interface {}{\"apiVersion\":\"actions.summerwind.dev/v1alpha1\", \"blockOwnerDeletion\":true, \"controller\":true, \"kind\":\"RunnerSet\", \"name\":\"example-runnerset\", \"uid\":\"e26f7d01-3168-496d-931b-8e6f97b776ea\"}}, \"uid\":\"4ee490f5-9a8c-4f30-86f9-61dea799b972\"}, \"spec\":map[string]interface {}{\"env\":interface {}(nil), \"image\":\"\", \"repository\":\"mumoshu/actions-runner-controller-ci\"}}: validation failure list:\nspec.env in body must be of type array: \"null\""}
github.com/go-logr/zapr.(*zapLogger).Error
```
and while the runnerdeployment controller is trying to create/update runners.
I've fixed it so that the new `RunnerDeployment` example added to README just works.
Removes an unnecessary condition from the deployment controller code. We assumed that the client would return a not-found error on an empty runnerset list it is clearly not the case.
Adds the initial version of RunnerDeployment that is intended to manage RunnerSets(#1), like Deployment manages ReplicaSets.
This is the initial version and therefore is bare bone. The only update strategy it supports is `Recreate`, which recreates the underlying RunnerSet when the runner template changes. I'd like to add `RollingUpdate` strategy once this is merged.
This depends on #1 so the diff contains that of #1, too. Please see only the latest commit for review.
Also see https://github.com/mumoshu/actions-runner-controller-ci/runs/471329823?check_suite_focus=true to confirm that `make tests` is passing after changes made in this commit.
RunnerSet is basically ReplicaSet for Runners.
It is responsible for maintaining number of runners to match the desired one. That is, it creates missing runners from `.Spec.Template` and deletes redundant runners.
Similar to ReplicaSet, this does not support rolling update of runners on its own. We might want to later add `RunnerDeployment` for that. But that's another story.
Yusuke Kuoka