Doesn't Create Volumes about chef-gluster HOT 10 CLOSED

Is there any error output?

I am guessing its because default['gluster']['server']['brick_mount_path'] already exists (/ is always used already so you can't mount something over the root partition)

from chef-gluster.

Hi,

There is no error. Everything runs ok and I don't see anything in the output where it is trying to create the volumes. On the server the peer probe is successful and the folders are created on the mount path it's just the final step of creating the volumes is not done.

To make things a little easier I went back to running this in test kitchen with your own kitchen.yml file settings.

Recipe:

#
# Cookbook Name:: mf-gluster
# Recipe:: server
#
# Copyright (c) 2015 The Authors, All Rights Reserved.

# Add the EPEL repo
yum_repository 'epel' do
  description 'Extra Packages for Enterprise Linux'
  mirrorlist 'http://mirrors.fedoraproject.org/mirrorlist?repo=epel-6&arch=$basearch'
  gpgkey 'http://dl.fedoraproject.org/pub/epel/RPM-GPG-KEY-EPEL-6'
  action :create
end

# Build server
include_recipe 'gluster::server'

Kitchen file:

- name: replicated1
    driver:
      ipaddress: 192.168.10.10
      vagrantfile_erb: test/integration/Vagrantfile.erb
      vm_hostname: gluster1
      additional_disk:
        size: 1
    run_list:
      - recipe[mf-gluster::set_hostname]
      - recipe[mf-gluster::server]
    attributes:
      gluster:
        server:
          brick_mount_path: "/data"
          disks:
            - sdb
          volumes:
            gv0:
              peers:
                - "gluster2"
                - "gluster1"
              replica_count: 2
              volume_type: "replicated"

  - name: replicated2
    driver:
      ipaddress: 192.168.10.20
      vagrantfile_erb: test/integration/Vagrantfile.erb
      vm_hostname: gluster2
      additional_disk:
        size: 1
    run_list:
      - recipe[mf-gluster::set_hostname]
      - recipe[mf-gluster::server]
    attributes:
      gluster:
        server:
          disks:
            - sdb
          brick_mount_path: "/data"
          volumes:
            gv0:
              peers:
                - "gluster2"
                - "gluster1"
              replica_count: 2
              volume_type: "replicated"

This is the output (I've only included the last bit of the master converge:

       Recipe: gluster::server_install

           - install version 3.1.1-16.el6 of package xfsprogs

           - install version 3.7.6-1.el6 of package glusterfs-server
        (up to date)

           - start service service[glusterd]
        (skipped due to only_if)
         * service[glusterfsd] action start (skipped due to only_if)
       Recipe: gluster::server_setup

           - execute "bash"  "/tmp/chef-script20151127-2451-81cqip"
         * execute[format partition] action run
           - execute mkfs.xfs -i size=512 /dev/sdb1
         * directory[/data/sdb1] action create



           - mount /dev/sdb1 to /data/sdb1

           - enable /dev/sdb1


           - restore selinux security context

           - execute gluster peer probe gluster1
        (skipped due to not_if)

       Running handlers:
       Running handlers complete
       Chef Client finished, 22/23 resources updated in 38.111062814 seconds
       Finished converging <replicated2-centos-66> (2m54.64s).
-----> Kitchen is finished. (7m14.89s)
zlib(finalizer): the stream was freed prematurely.
zlib(finalizer): the stream was freed prematurely.
iang@matches:~/repos/matches/chef$ kitchen login replicated1
/opt/chefdk/embedded/lib/ruby/gems/2.1.0/gems/httpclient-2.6.0.1/lib/httpclient/webagent-cookie.rb:458: warning: already initialized constant HTTPClient::CookieManager
/opt/chefdk/embedded/lib/ruby/gems/2.1.0/gems/httpclient-2.6.0.1/lib/httpclient/cookie.rb:8: warning: previous definition of CookieManager was here
Last login: Fri Nov 27 12:46:45 2015 from 10.0.2.2
[vagrant@gluster1 ~]$ ll /data/
total 0
drwxr-xr-x. 3 root root 16 Nov 27 12:47 sdb1
[vagrant@gluster1 ~]$ ll /data/sdb1/gv0/
total 0
[vagrant@gluster1 ~]$ sudo -i
[root@gluster1 ~]# gluster peer status
Number of Peers: 1

Hostname: gluster2
Uuid: 24782800-50b4-453d-be6b-859b28d4b4ec
State: Peer in Cluster (Connected)
[root@gluster1 ~]# gluster volume status
No volumes present

Thanks for your help

from chef-gluster.

The LVM volumes code is broken. Use the disks attribute instead of lvm_volumes please for now, until the latest version is released.

from chef-gluster.

Great thanks. I ended up writing something myself to create the replicas
for us.

Thanks
Ian

On Thu, Dec 3, 2015 at 2:59 PM, Andrew Repton [email protected]
wrote:

The LVM volumes code is broken. Use the disks attribute instead of
lvm_volumes please for now, until the latest version is released.

—
Reply to this email directly or view it on GitHub
#49 (comment)
.

from chef-gluster.

Hi All

I used the disks attributes and configure a 2 two node gluster with the configuration below.

node.default['gluster']['server']['brick_mount_path'] = "/data"
node.default['gluster']['server']['disks'] = ['sdb']
node.default['gluster']['server']['volumes'] = {
'gv1' => {
'peers' => ['node1', 'node2'],
'replica_count' => 2,
'volume_type' => "replicated"
}
}

The peer probe is working but I get these messages:
-Correct number of bricks not available: 1 needs to be at least 1. Skipping...
-I can't invite myself into the pool!

I created it manually without an issue to confirm the peers are working.
-gluster volume create gv1 replica 2 node1:/data/sdb1/gv1 node2:/data/sdb1/gv1

I clone the repo and ran the default kitchen setup and had the same issue.

Are there any know issues beside LVM or am I doing something wrong?

Thanks,
Andrew

from chef-gluster.

Hi @acholt

Your chef lookup is failing. The primary node cannot find the other node2 in order to confirm the brick is there. Please confirm the node name is the same as the FQDN (the current code requires this). @shortdudey123 and I are working on a fix for this right now.

Thanks!

Andy

from chef-gluster.

Hay @Seth-Karlo

Thanks for the reply. :-)

I checked the node name and the FQDN are the same. I also tried adding 'peer_names' => ['node1','node2'], to my config but it made no difference.

[2015-12-07T16:18:18+00:00] WARN: Unable to find a chef node for node2 [2015-12-07T16:18:18+00:00] WARN: Correct number of bricks not available: 1 needs to be at least 1. Skipping...
[2015-12-07T16:18:18+00:00] WARN: I can't invite myself into the pool!
Converging 13 resources
Recipe: testcookbook::set_hostname
- execute echo 172.28.128.50 node1 >> /etc/hosts
- execute echo 172.28.128.51 node2 >> /etc/hosts

[root@node1 ~]# gluster peer status
Number of Peers: 1

Hostname: node2
Port: 24007
Uuid: 995acb5b-a166-4f38-b823-997ba97a01be
State: Peer in Cluster (Connected)
[root@node1 ~]# ohai fqdn
[
"node1"
]

[root@node2 ~]# gluster peer status
Number of Peers: 1

Hostname: node1
Port: 24007
Uuid: 070f82a8-ab1f-41a6-93cb-0c54915a8737
State: Peer in Cluster (Connected)
[root@node2 ~]# ohai fqdn
[
"node2"
]

Many Thanks,
Andrew

from chef-gluster.

@acholt are you still having the issue? if so, can you try the newest version and see if the issue still occurs?

from chef-gluster.

Closing due to inactivity
The cookbook has changed quite a bit since this issue was opened
Please create a new issue if this is still an issue with the current version

from chef-gluster.

@shortdudey123 apologizes for not responding sooner. I have tested it thoroughly and all is working as expected. :-)

Many Thanks,
Andrew

from chef-gluster.