Ask Your Question

help with failing Mirantis fuel 9.1 deployment

asked 2017-01-25 11:12:11 -0500

racingferret gravatar image

updated 2017-01-25 11:12:20 -0500

Hi All,

I'm currently trying to deploy a basic 4 node PoC with Mirantis fuel 9.1 All hardware is identical and the comms check passes on the Fuel dashboard. I have one controller node and 3 compute/osd nodes. Cinder, Glance, Nove and Swift are all backed by Ceph with an object replication factor of 3.

The controller node finishes deployment without issue, but all the compute/osd nodes fail as they timeout running ceph commands. On the controller node, if I run "ceph -s", I get the following output:

cluster a7f64266-0894-4f1e-a635-d0aeaca0e993
  health HEALTH_ERR 192 pgs stuck inactive; 192 pgs stuck unclean; no osds
  monmap e1: 1 mons at {node1=}, election epoch 1, quorum 0 node1
  osdmap e1: 0 osds: 0 up, 0 in
  pgmap v2: 192 pgs, 3 pools, 0 bytes data, 0 objects
     0 kB used, 0 kB / 0 kB avail
     192 creating

The ceph log file on and controller node also outputs the following:

mon.node-14@0(leader).auth v33 caught error when trying to handle auth request, probably malformed request

Running any ceph command on any of the osd nodes results in a pause, followed by the following being output continuously:

monclient: hunting for new mon

Any help with how to start debudding this would be great.


edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted

answered 2017-02-15 18:17:12 -0500

racingferret gravatar image

In case anyone else stumbles across this issue, turns out the non-default MTU of 9000 I set for the storage network wasn't being set on the vswitch. Setting it back to "default" allowed the install to complete without issue.

edit flag offensive delete link more

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower


Asked: 2017-01-25 11:12:11 -0500

Seen: 315 times

Last updated: Feb 15 '17