gce_vm_cluster function

Make a VM cluster suitable for running parallel workloads

This wraps the commands for creating a cluster suitable for future workloads.

gce_vm_cluster(vm_prefix = "r-cluster-", cluster_size = 3,
  docker_image = "rocker/r-parallel", ..., ssh_args = NULL,
  project = gce_get_global_project(), zone = gce_get_global_zone())

Arguments

• vm_prefix: The prefix of the VMs you want to make. Will be appended the cluster number
• cluster_size: The number of VMs in your cluster
• docker_image: The docker image the jobs on the cluster will run on. Recommend this is derived from rocker/r-parallel
• ...: Passed to gce_vm_template
• ssh_args: A list of optional arguments that will be passed to gce_ssh_setup
• project: The project to launch the cluster in
• zone: The zone to launch the cluster in

Examples

## Not run:

library(future)
library(googleComputeEngineR)

vms <- gce_vm_cluster()

## make a future cluster
plan(cluster, workers = as.cluster(vms))
## End(Not run)