Created attachment 1175101 [details]
Hosted Engine setup log

Description of problem:
The HE deployment gets stuck if the glusterfs version on the shared storage host is different from the version on the node. According to Sahina Bose @ RedHat [1], this is a compatibility issue on executing cli commands. --remote-host option throws errors when there's a version mismatch. The HE deployment should handle this error so that installation can proceed.

[1] http://lists.ovirt.org/pipermail/users/2016-July/040814.html

How reproducible:
Always

Steps to Reproduce:
1. Have a gluster server running a version of glusterfs not equal to 3.7.11.
2. Install a node using ovirt-node-ng-installer-ovirt-4.0-2016062412 ISO.
3. Go to https://<node>:9090/ and start HE deployment, or launch using ovirt-hosted-engine-setup.
4. Choose glusterfs as the shared storage.
5. At the prompt "Please specify the full shared storage connection path to use", enter a path to a gluster volume on the server in step 1.

Actual results:
The prompt will repeat. In the Cockpit portal, it returns a False value in the field. In the console, it will give the following error: [ ERROR ] Cannot access storage connection storage1.nullaxiom.com:/mgmttank: Command '/sbin/gluster' failed to execute

Expected results:
The installation should continue.

Additional info:
The setup log, console output, and cli.log are attached.