Hello,
In our organization we have various teams with exceptionally large repositories. This can cause longer deploy times when a Harness Delegate must re-fetch an entire repository. In some cases, it can cause unexpected failures if the dev team has not set the timeout for their deployment high enough.
In most cases, a full repository checkout is not necessary (i.e. pulling helm overrides). It would be faster to perform a sparse checkout. We understand a sparse checkout is possible by creating a pre-deployment step and then executing the deploy on the same delegate. However, this is not preferred as this creates a single point of failure on the single Harness Delegate. We consider it best practice to not use delegate selectors wherever possible such that any delegate can be selected for a job.
With that background, we would like to request that Harness support a sparse checkout directly in the deploy steps such that the sparse checkout does not limit the deployment to a single specific Harness delegate.