2

I have set up a anaconda cluster with AWS ec2 following the guide on anaconda official website http://docs.continuum.io/anaconda-cluster/quickstart. Everything works perfect.

But if I stop the ec2 machine used by acluster manually and then restart that machine, everything broken, because the machine starts with another public ip address.

type in 

acluster list clusters

can see the ip address of compute/head machines are still holding old ip address. Is there a way to update the machine ip address of the cluster AUTOMATICALLY? or convenient way to handle this?

mattexx
  • 6,456
  • 3
  • 36
  • 47
Hello lad
  • 17,344
  • 46
  • 127
  • 200

1 Answers1

0

Right now, Anaconda Cluster does not support automatic updates of the IP addresses when you restart the EC2 instances. As we improve the library, we are considering handling reboots and automatic updating of IP addresses.

As a workaround, you can edit the ~/.acluster/clusters.d/your_cluster.yaml file and update the IP addresses manually.

koverholt
  • 26
  • 4
  • thanks a lot for the response. After update the IP addresses manually, the "acluster open notebook" doesn't work anymore – Hello lad Jan 15 '16 at 15:24
  • If you installed Jupyter notebook via `acluster install notebook`, then the notebook server should be running on the head node, which is in the `~/.acluster/clusters.d/your_cluster.yaml` under the machines > head section. – koverholt Jan 25 '16 at 19:26