gluster / gluster-ansible-features

Ansible role to enable a Gluster cluster with gluster-specific features like samba, ganesha etc.

gluster.features

The gluster.features role implements GlusterFS usecases. This role implements following sub-roles:

• nfs_ganesha
• gluster_hc
• ctdb
• geo_replication

Requirements

• GlusterFS

Role Variables

NFS Ganesha related variables

Name	Choices	Default value	Comments
gluster_features_ganesha_haname		UNDEF	Name of the NFS Ganesha cluster.
gluster_features_ganesha_volume		UNDEF	An existing GlusterFS volume which will be exported through NFS Ganesha
gluster_features_ganesha_hostnames		UNDEF	A comma separated list of hostnames, these are subset of nodes of the Gluster Trusted Pool that form the ganesha HA cluster
gluster_features_ganesha_viplist	UNDEF	public	A comma separated list of virtual IPs for each of the nodes specified above.
gluster_features_ganesha_masternode		UNDEF	One of the nodes from the Trusted Storage Pool, gluster commands will be run on this node. gluster_features_ganesha_masternode: "{{ groups['ganesha_nodes'][0] }}" - the first node of the inventory section ganesha_nodes will be used.
gluster_features_ganesha_clusternodes		UNDEF	List of the nodes in the Trusted Storage Pool. gluster_features_ganesha_clusternodes: "{{ groups['ganesha_nodes'] }}" - The nodes listed in section ganesha_nodes in the inventory.

Gluster Hyperconverged Interface setup related roles

Name	Choices	Default value	Comments
gluster_features_hci_cluster		UNDEF	The cluster ip/hostnames. Can be set by gluster_hci_cluster: "{{ groups['hc-nodes'] }}", where hc-nodes is from the inventory file.
gluster_features_hci_volumes		UNDEF	This is a dictionary setting the volume information. See below for further explanation and variables.
gluster_features_hci_volume_options		UNDEF	This is not needed to be set by user, defaults are picked up. Set to override defaults. For default values see Gluster HCI documentation.
gluster_features_hci_master		UNDEF	The REST host to be connected to (do not include `http'). This value will be ignored if glusterd1 is running on remote node.
gluster_features_hci_port		24007	The glusterd2 port

Dependencies

gluster.infra gluster.cluster

Example Playbook

Creating a NFS Ganesha Cluster:

An example playbook to deploy NFS Ganesha. Note to gather_facts should be set to true.

---
- name: Setting NFS Ganesha
  remote_user: root
  hosts: ganesha_nodes
  gather_facts: false

  vars:
     # Setting up NFS Ganesha
     gluster_features_ganesha_haname: ganesha-ha
     gluster_features_ganesha_volume: nfs_ganesha
     gluster_features_ganesha_hostnames: "server1,server2,server3,server4"
     gluster_features_ganesha_viplist: 10.70.44.121,10.70.44.122,10.70.44.123,10.70.44.124

     gluster_features_ganesha_masternode: "{{ groups['ganesha_nodes'][0] }}"
     gluster_features_ganesha_clusternodes: "{{ groups['ganesha_nodes'] }}"

  roles:
     - gluster.features

The above playbook assumes that a volume named nfs_ganesha is created and running.

Note:

If you are using GlusterFS 4.0 (glusterd2) ensure to install python-gluster-mgmt-client package on the remote master host.

Or create a directory glusterapilib on the remote host and copy the files volume.py, peer.py, common.py,init.py, excpetions.py under glusterapilib in a python searchable path.

License

GPLv3

Author Information

Sachidananda Urs [email protected]

gluster-ansible-features's Issues

[CTDB] adding local user via smbpasswd fails on Ubuntu 20.04

When the flow is reaching the point to create the user, smbpasswd fails with :

TASK [gluster.features/roles/ctdb : Set smb password] *****************************************************************************************************************************************
fatal: [kubluster]: FAILED! => {"changed": true, "msg": "non-zero return code", "rc": 1, "stderr": "Shared connection to kubluster closed.\r\n", "stderr_lines": ["Shared connection to kubluster closed."], "stdout": "\r\n/bin/sh: 1: 1.1: not found\r\n/bin/sh: 1: AES256: not found\r\n/bin/sh: 1: 1.1: not found\r\n/bin/sh: 1: AES256: not found\r\nFailed to open /var/lib/samba/private/secrets.tdb\r\n", "stdout_lines": ["", "/bin/sh: 1: 1.1: not found", "/bin/sh: 1: AES256: not found", "/bin/sh: 1: 1.1: not found", "/bin/sh: 1: AES256: not found", "Failed to open /var/lib/samba/private/secrets.tdb"]}

It seems that 'clustering=yes' is the root cause. Commenting out the line allows the change of the pass.

ctdb role hangs while trying to stop gluster volume

Ansible playbook execution hangs at the following stage:

TASK [gluster.features/roles/ctdb : Add volume name to ctdb setup/teardown scripts] ***
skipping: [storage0] => (item=/var/lib/glusterd/hooks/1/stop/pre/S29CTDB-teardown.sh) 
skipping: [storage0] => (item=/var/lib/glusterd/hooks/1/start/post/S29CTDBsetup.sh) 
skipping: [storage1] => (item=/var/lib/glusterd/hooks/1/stop/pre/S29CTDB-teardown.sh) 
skipping: [storage1] => (item=/var/lib/glusterd/hooks/1/start/post/S29CTDBsetup.sh) 

TASK [gluster.features/roles/ctdb : Stop GlusterFS volume] *********************

Looking at the particular task it seems like gluster volume stop <VOLNAME> force is waiting for an input for the prompt Stopping volume will make its data inaccessible. Do you want to continue? (y/n). Using --mode=script to the above command is a viable solution.

oVirt single node hyperconverged fails after changing gluster commands to CLI

After #54 setting up a single node fails on the following steps

Create the GlusterFS volumes
Set virt to gluster volume

In my testing changing "Create the GlusterFS volumes" to only use the replica option if it's defined and skipping the "Set virt" step for a single host allows the set up to finish

installing nfs ganesha does not work.

Given the following playbook:

- name: install ganesha
  hosts: ganesha
  become: true
  vars:
    gluster_features_ganesha_volume: vol1
    gluster_features_ganesha_hostnames: 
    - "{{ groups['ganesha'] }}"
    gluster_features_ganesha_haname: ganesha
    gluster_features_ganesha_viplis: "192.168.21.21,192.168.21.22"
    gluster_features_ganesha_clusternodes:
    - "{{ groups['ganesha'] }}"
    gluster_features_ganesha_masternode: "{{ groups['ganesha'][0] }}"

  roles:
  - gluster.features

This fails at the "Copy the ssh keys to the local machine" task.

The reason is that it attempts to do this before those keys have been generated.

Add tasks/check_vars.yml for every role

This tasks file should essential check for the required variables for that particular role.

For example:

- name: Checking that required variables are set
  fail: msg="{{ item }} is not defined"
  when: not {{ item }}
  with_items:
    - foo_volname
    - foo_username
    - foo_mastenodes

This should essentially do a role sanity check before executing the role.

nfs_ganesha: Get rid of variable gluster_features_ganesha_masternode

As of now the variable gluster_features_ganesha_masternode provides the user flexibility to choose the master node, at the cost of extra pain in assigning the variable.

We can always choose the first node as master node and get rid of this variable.