0

Given is an OpenShift Origin cluster of four nodes. Previously, an containerized GlusterFS cluster has been added, distributed over three nodes, with node 4 not being used for GlusterFS. Initial installation and GlusterFS installation has been done with openshift-ansible.

Now I would like to add the fourth node to the GlusterFS cluster, but I haven't found a proper way to do it. I assume that just adding a node with Heketi won't be enough, while re-running GlusterFS playbooks is not recommended. How can I still use my fourth node as additional node in GlusterFS cluster?

Graham Dumpleton
  • 57,726
  • 6
  • 119
  • 134
Nikolai Prokoschenko
  • 8,465
  • 11
  • 58
  • 97

1 Answers1

0

You need to follow this section:

12.1.3.1. Adding a New Cluster to the Existing Container-Native Storage

from here: https://access.redhat.com/documentation/en-us/red_hat_gluster_storage/3.3/html-single/container-native_storage_for_openshift_container_platform/#idm139668586251520

In short:

  1. you need to label the nodes.

oc label node

  1. load the topology with new node like

heketi-cli topology load..

Go through the entire section of 12.1 to get a overview of how to increase the capacity.

kumar
  • 2,530
  • 6
  • 33
  • 57
  • Are you sure that's everything that needs to be done? From what I see, glusterfs pods mount a lot of directories from the host system which probably needs to be initialized somehow? – Nikolai Prokoschenko Sep 13 '18 at 10:13