Reduxio Puppet module for managing Reduxio HX Series, using a network device or locally. The module enables the user to create, modify and update volumes, hosts and assignment.
Note: This module does not automate the configuration of the iSCSI initiator. This includes the configuration of iSCSI sessions, but also the Reduxio iSCSI best practices documented in the Reduxio Support Portal. These should be configured manually or using another Puppet module.
- Reduxio HX Series system running Reduxio TimeOS v3.3 or higher.
- 'rest-client' Ruby Gem.
- Puppet v3.6 or higher.
shell> puppet module install rdxdev-reduxio
Note: It is recommended to restart puppet and puppetmaster services after module installation.
Reduxio Puppet module can be used either as a network device or locally.
reduxio_history_policy { 'Default-Apps':
ensure => 'present',
is_default => 'true',
seconds => 86400,
hours => 336,
days => 14,
weeks => 12,
months => 12,
retention => 12,
}
# The volume title is also the name of the volume that will be created on the Reduxio system.
reduxio_volume { 'vol1':
ensure => 'present',
description => 'Volumes creation with puppet',
size => 11,
history_policy => 'Critical-Apps',
blocksize => 512
}
reduxio_volume { 'vol2':
ensure => 'present',
description => 'Volume creation with puppet',
size => 101,
history_policy => 'Default-Apps',
blocksize => 512
}
reduxio_volume { 'vol3':
ensure => 'present',
description => 'Volume creation with puppet',
size => 100,
history_policy => 'Bookmarks-Only',
blocksize => 4096
}
reduxio_hg { 'hg1':
ensure => 'present',
description => 'Hostgroup created by puppet',
}
reduxio_host { 'host':
ensure => 'present',
# Omitting 'iscsi_name' field will create a host with iscsi_name as the agent that is executing the manifest.
# Using this feature relies on the /etc/iscsi/initiatorname.iscsi file.
iscsi_name => 'iqn.2010-10.example'
hg_id => 'hg1'
user_chap => 'host_chap_user',
password_chap => 'host_chap_password',
description => 'host description',
}
# The most recommended way to define a volume to host assignment is in the entity title: <vol_name>/<host_name>.
reduxio_volume_to_host_assignment {'vol1/host':
ensure => 'present'
}
# Alternative way to define a volume to host assignment.
# The title in this case will not have any functional effect on the Reduxio assignment.
reduxio_volume_to_host_assignment {'second_host_assignment':
ensure => 'present',
volume => 'vol2',
host => 'host'
}
# The most recommended way to define a volume to host assignment is in the entity title: <vol_name>/<hostgroup_name>.
reduxio_volume_to_hg_assignment {'vol3/hg1':
ensure => 'present'
}
# Alternative way to define a volume to host group assignment.
# The title in this case will not have any functional effect on the Reduxio assignment.
reduxio_volume_to_hg_assignment {'second_hg_assignment':
ensure => 'present',
volume => 'vol3',
host => 'host'
}
# Recommended entities execution order
Reduxio_history_policy <| |> -> Reduxio_volume <| |> -> Reduxio_hg <| |> -> Reduxio_host <| |> -> Reduxio_volume_to_host_assignment <| |> -> Reduxio_volume_to_hg_assignment <| |>
Puppet plugin is using a Reduxio generated API Token. To create a token, open this URL: https://reduxio_system_address/#/app/settings/api-tokens.
Updating objects' attributes is supported, however as names are the unique identifiers of objects, changing objects names is not possible.
The Reduxio Puppet module can act as a network device. Example device.conf file, which should reside in a node which shall
be the proxy to the Reduxio system (aka proxy node):
[devicename]
type reduxio
url https://api-token@reduxio_system_address
The device name (or cert name) is unrelated to the agent node host name. Furthermore, it is recommended not to name the devices as the proxy node host name, to avoid conflict in cases where the proxy node should act both as agent and as a Reduxio network device(s) proxy. The device name is the certificate name that the puppet master (server) will be authenticated with. If working with the same Reduxio system from multiple proxy nodes, make sure to name the devices in each node differently, if wishing to execute different manifests from the proxy nodes on the same Reduxio system.
Once defining a device.conf file on the proxy node, you should also define the relative manifest to the network device
on the puppet master node (usually in site.pp file):
node 'devicename' {
reduxio_volume { 'vol1':
ensure => 'present',
description => 'Volumes creation with puppet',
size => 11,
history_policy => 'Critical-Apps',
blocksize => 512
}
}
To dispatch the network device configuration, run the following on the proxy node:
shell> puppet device
Note: If the network device hasn't been authenticated with the puppet master, you should specify --waitforcert
parameter and sign the request on the puppet master while the proxy node is waiting for authentication.
After signing, you should run again the puppet device command again.
If using a non-default device.cong file, you can specify an alternative conf file:
shell> puppet device --deviceconfig /path/to/alternative/conf/file
Using the Reduxio Puppet module locally can be achieved by explicitly defining a 'url' attribute per entity, allowing the management of multiple Reduxio machines in a single manifest file. For example:
reduxio_volume { 'vol1':
ensure => 'present',
description => 'Volumes creation with puppet',
size => 11,
history_policy => 'Critical-Apps',
blocksize => 512,
url => 'https://api-token1@reduxio1'
}
reduxio_volume { 'vol2':
ensure => 'present',
description => 'Volume creation with puppet',
size => 101,
history_policy => 'Critical-Apps',
blocksize => 512,
url => 'https://api-token2@reduxio2'
}
This will create two volumes: vol1 on reduxio1 machine, and vol2 on reduxio2 machine. When working with multiple
machines, a title is unique per entity type, so one needs to explicitly define the name property when multiple entities
with the same name should be created across multiple Reduxio systems.
This approach can also be combined with the network device approach: Specifying the url attribute will override the
device.conf settings.