GitHub is home to over 20 million developers working together to host and review code, manage projects, and build software together.
When a cluster is started with dns_prefix=True, the defaults in get/put still only look for master, not <dns_prefix>-master.
i.e. you get
$starcluster get mycluster /path/to/my/file .
StarCluster - (http://star.mit.edu/cluster) (v. 0.95.3)
!!! ERROR - node 'master' does not exist
Which can be easily handled with
$starcluster get mycluster -n mycluster-master /path/to/my/file .
but this fix is simple enough that I thought I'd send it.
Fix get/put when using dns_prefix
Wow, I am so less likely to contribute changes back to sc.
Obvious change is obvious
Done fixing pep-8 nonsense
This is because we don't load the metadata when fetching the cluster in the put/get commands (ie load_receipt=False). If we flip that to True then dns_prefix will be toggled and we can skip duplicating the dns_prefix node fetch logic. I'd probably also pass load_plugins=False, and load_volumes=False when fetching the cluster since they're not used by put/get.