Dynamic Worker Pools are a special type of worker pool used by our cloud product to intialize a group of workers, when a task is assigned to a worker, the task will be executed by one of the workers in the pool.
The Default Worker Pools for our latest cloud instances are configured to use a Dynamic Worker Pool, which means it displays some different characteristics to older cloud instances and our self-hosted product.
Workers are created on demand and are assigned to a customer when required. Once you've finished using a worker, the worker is destroyed and not reused.
Octopus Cloud defines the use of the worker as finished when it has been idle for an hour (60 minutes). If your cloud instance does another deployment or runbook run within an hour, the same worker will be re-used. No matter what, Octopus Cloud will destroy the worker after 72 hours (3 days). These settings can be adjusted for your instance. Reach out to email@example.com to request any adjustments.
Each worker is provisioned so that they provide your task a sandbox that is completely isolated from our other customers.
Customers cannot make changes to the default worker pool, or the configuration of the workers themselves. The workers are provisioned using our own service, meaning you don't need to configure anything for them to work in your environment.
Each worker is provisioned with a baseline of tools including, but not limited to:
- Octopus Client (latest)
- .NET Frameworks (3.5+)
- Chocolatey (latest)
- Nuget CLI (latest)
- AWS IAM Authenticator (0.3.0)
- Helm (2.9.1)
- Python & pip (3.7.4)
- Microsoft Service Fabric SDK (3.0.480)
- Microsoft Service Fabric (6.1.480.9494)
- Kubectl (1.14.9)
The version of
kubectl can be overridden by setting the variable
Octopus.Action.Kubernetes.CustomKubectlExecutable to the following value
where version is one of the following:
The versions and tools used above are subject to change, this list is intended to represent the general configuration at the time of writing.
Some of the tool installations are locked to a specific version, where others may be running the latest version.
Need support? We're here to help.