| We're seeing a number of failures on ci.jenkins.io which stem from Azure Container Agents failing to provision. It looks like failed ACI deployments don't get cleaned up automatically by the plugin. ACI is returning the following failure:
{"code":"DeploymentFailed","message":"At least one resource deployment operation failed. Please list deployment operations for details. Please see https://aka.ms/arm-debug for usage details.","details":[{"code":"Conflict","message":"{\r\n \"error\": {\r\n \"code\": \"ServiceUnavailable\",\r\n \"message\": \"The requested resource is not available in the location 'eastus2' at this moment. Please retry with a different resource request or in another location. Resource requested: '2' CPU '4' GB memory 'Linux' OS\"\r\n }\r\n}"}]}
Which is annoying, but whatever. The bug here is that the plugin does not clean up this failed deployment, so we have exceeded our quota of Deployments of 800, all failures like this one! I recall this exact same issue in the Azure VM Agents plugin at one point as well.  |