When you try to do this you will encounter the following error: You cannot queue an Azure Pipelines job on an empty agent pool because Azure Pipelines cannot validate if the pool matches the requirements for the job. PoolID: "1" organizationURLFromEnv: "AZP_URL" personalAccessTokenFromEnv: "AZP_TOKEN" Placeholder agent PollingInterval: 30 successfulJobsHistoryLimit: 5 failedJobsHistoryLimit: 5 maxReplicaCount: 10 scalingStrategy: Running Azure Pipelines jobsĪfter deploying the agent and the KEDA ScaledObject it is time to see the autoscaling in action.įirst, check the current pods running in the deployment: ⚠ The Azure Pipelines scaler supports scaling to zero but you need at least one agent registered in the agent pool in order for new jobs to be scheduled on the pool. The default targetPipelinesQueueLength is 1, so there will be one agent for each job. PoolID: "1" organizationURLFromEnv: "AZP_URL" authenticationRef: MinReplicaCount: 1 maxReplicaCount: 5 triggers: You can easily deploy the agent as a Kubernetes deployment by using this Kubernetes manifest: To create a basic Azure Pipelines agent image you can follow the instructions from the official docs. Deploying a self-hosted agent on Kubernetes as a Deployment Create the container image You can run the agents as a Deployment or a Job in Kubernetes and scale them accordingly with a ScaledObject or a ScaledJob. However, with KEDA v2.3 you can now autoscale your self-hosted agents on Kubernetes based on the amount of pending jobs in your agent pool. When running self-hosted agents on Kubernetes, there is no out-of-the-box support for autoscaling.
Pool queue full#
But if you want full control, you are going to have to use self-hosted agents.Īgents are able to run on Linux, macOS or Windows machines and can be packaged in to a container. What are Azure Pipelines self-hosted agents?Īzure Pipelines jobs can run on different kinds of agents ( docs ). KEDA will autoscale to the maximum defined in the ScaledObject and does not limit itself to the parallel jobs count defined for the Azure DevOps organization. 💡 The number of concurrent pipelines you can run is limited by your parallel jobs. By autoscaling the agents you can create a scalable CI/CD environment. Self-hosted Azure Pipelines agents are the perfect workload for this scaler. With the addition of Azure Piplines support in KEDA, it is now possible to autoscale your Azure Pipelines agents based on the agent pool queue length.
0 kommentar(er)
