The repository contains StorPool's playbook and roles for Ansible.
# Running this playbook with ansible-playbook:
ansible-playbook /home/tools/ansible/playbook.yml [parameters]
# This playbook can be also execuded directly from the playbook's directory, like this:
/home/tools/ansible/playbook.yml [parameters]
| Parameter (Long) | Parameter (Short) | Parameter Description |
|---|---|---|
| --inventory | -i | specify inventory host path or comma separated host list |
| --limit | -l | further limit selected hosts to an additional pattern |
| --list-tags | list all available tags | |
| --tags | -t | only run plays and tasks tagged with these values |
| --skip-tags | only run plays and tasks whose tags do not match these values | |
| --step | one-step-at-a-time: confirm each task before running | |
| --flush-cache | clear the fact cache for every host in inventory |
# Copy the hosts.example file:
cp hosts.example ansible.hosts
# Edit the hosts file and configure StorPool parameters:
vi ansible.hosts
# Save the file and run the playbook:
ansible-playbook /home/tools/ansible/playbook.yml -i ansible.hosts
# Or run it directly:
/home/tools/ansible/playbook.yml -i ansible.hosts
| Tag | Description |
|---|---|
| prerequisites | Set common variables for target hosts, do some common tests |
| setup-infra | Install StorPool's dependencies on target hosts, OS related configurations and performance tuning tasks |
| install | Download and install StorPool and the StorPool support tools |
| setup-network | Configure and validate StorPool's network |
| setup-drives | Initialize StorPool drives |
| setup-cgroups | Configure cgroups as per StorPool's requirements |
| setup-services | Enable and start StorPool systemd services |
| reboot-hosts | Reboot target hosts |
| tests | Aftertests |
# Show list of available tags:
ansible-playbook /path/to/ansible/playbook.yml --inventory ansible.hosts --list-tags
# Run specific tags:
ansible-playbook /path/to/ansible/playbook.yml --inventory ansible.hosts --tags prerequisites
# Skip specific tags:
ansible-playbook /path/to/ansible/playbook.yml --inventory ansible.hosts --skip-tags setup-network,setup-drives
# Run specific task on a specific server from your inventory:
ansible-playbook /path/to/ansible/playbook.yml --inventory ansible.hosts --limit example-server.1 --tags setup-network
NOTE: We use fact caching in order to speed up the process by caching all required variables on the ansible server (inside
./factsdirectory) for later runs. If you have deleted the facts directory or any of the hosts facts inside it, you need to executeansible-playbook /path/to/ansible/playbook.yml --inventory ansible.hosts --tags variablesin order for you to re-gather all required variables inside the facts cache.
| Variable | Description |
|---|---|
| sp_cluster | This is the top directory where Ansible configuration is saved. Usually basename $(pwd) |
| sp_custdir | The parent directory of sp_cluster. This is dirname $(pwd) |
| sp_toolsdir | The path on the destination hosts where StorPool files will be stored Defaults to /root/storpool |
| sp_release_file | Full path the StorPool tar.gz archive, or web to download it from StorPool's repo Defaults to web |
| sp_release | StorPool version to be deployed (if sp_release_file='web') Defaults to 19.01 |
| sp_reinstall | Whether to always attempt to re-install regardless if a previous version was installed or not Defaults to True |
| sp_new_cluster | Deployment type Defaults to False |
| sp_multi_server_helper_enabled | Redistributes the disks on multiple server instances Defaults to True |
| sp_config | Path to storpool.conf on the ansible host Defaults to the inventory directory (Where the ansible inventory file is located) |
| sp_update_system | Perform OS update (yum/apt update/upgrade) on target hosts Defaults to True |
| sp_retry_install | Retry StorPool installation to handle transient network errors Defaults to False |
| sp_alternative_net_setup | Used for an alternative networking setup tooling during internal testing Defaults to False |
| sp_configure_network | Automatically configure network interfaces based on storpool.conf variables Defaults to False |
| sp_overwrite_iface_conf | Overwrite existing ineterface configuration files (iface-genconf -o). Defaults to False |
| sp_single_iface | Specifies --nettype 0 on iface_genconf Defaults to False |
| sp_selinux | Set SELinux state (valid options are "permissive" or "disabled") Defaults to disabled |
| sp_disable_nm | Stop and disable NetworkManager Defaults to True |
| sp_disable_fw | Stop and disable firewalld/ufw (if False, will add ports for StorPool) Defaults to True |
| sp_vm | Nodes are virtual machines - It will be determined automatically if not specifically set |
| sp_summary_wait | Generate a summary and pause after variable validation Defaults to True |
| sp_cg_hyperconverged | List of arguments to pass to storpool_cg conf |
| sp_cg_conf_extra | List of extra arguments to pass to storpool_cg conf |
| sp_diskid_offset | Offset to calculate diskid prefix (useful when sp_node_id >= 40) |
| sp_kernel_params_reboot | Automatically reboot to apply required kernel parameters set by install.sh Defaults to False |
For more information please check hosts.example file
# Gather all required playbook variables to cache:
ansible-playbook /path/to/ansible/playbook.yml --inventory ansible.hosts --tags prerequisites
# Flush facts cache and gather everything again:
ansible-playbook /path/to/ansible/playbook.yml --inventory ansible.hosts --tags prerequisites --flush-cache
- Auto-enable storpool_nvmed if nvme drives have been initialized
- Add integration roles (onapp, openstack, cloudstack opennebula, etc)
- Ubuntu: more testing to address specific cases