Clusterless Examples

Warning, much assembly required.

Here are a few examples using the Clusterless cli.

To learn more, see the documentation:

• User Guide
• Reference

Consider exploring them in order:

• native-copy
• native-copy-chain
• python-batch

Notes

The examples rely on jsonnet to generate the JSON files. This is not a requirement imposed by Clusterless but a convenience to creating simple project files that don't need to be edited before deployment. e.g. S3 bucket names are global, so need to be unique to the deployment.

See 0002-static-json-as-configuration.md.

The default boundary has eventArrival: "infrequent", this means the boundary will only be triggered once per upload per interval. Two uploads in the same interval will cause an error (error management here is a TODO).

You can change this to eventArrival: "frequent" to have the boundary accumulate events within an interval. This has a slightly higher cost as the boundary polls for events at the end of every interval.

Prerequisites

To install see:

• https://github.com/ClusterlessHQ/clusterless#installing-clusterless

You should also install these tools:

• jsonnet
• jq

Bootstrap

After installing the AWS CDK for a given region, configure the region below:

Copy .properties.libsonnet.template to .properties.libsonnet and edit.

Then bootstrap Clusterless into the region:

jsonnet bootstrap.jsonnet | jq -r '.[0]'

Copy/paste the bootstrap command to run.

Current limitations

Clusterless is under active development. There are bugs, please post them.

• https://github.com/ClusterlessHQ/clusterless/issues

See the Roadmap.