Clusters created does not automatically clean up

[KangChungYang]

Here is my code:

cl <- create_cluster(8)
set_default_cluster(cl)
data_par <- data %>% partition(IID, cluster=cl)
cluster_assign_value(cluster=cl, "myfun", myfun)
out <- data_par %>% do(res=myfun(.)) %>% collect()

After that, sometimes the clusters created does not automatically clean up Is there any function can close those clusters?

[Ax3man]

data_par is still partitioned, so the cluster should not be cleaned up, I think. You can use parallel::stopCluster(cl) to close it anyway.

[hadley]

You need to delete data_par and then the gc will eventually clean up the cluster.