Summary¶

This documentation provides both simple and advanced views into running the ScaleIO Puppet module. There should be enough documentation here to solve for most viable use cases.

EMC CODE State¶

• Experimental

Get ScaleIO RPMs¶

In order to perform the install of the components on the VMs, you must have the ScaleIO RPMs. These can be downloaded directly from the support.emc.com website.

Update ScaleIO Version¶

There is one parameter that must be modified depending on your ScaleIO RPM version numbers. There is a global parameter in the puppet/manifests/site.pp file that needs to be updated.

Check your RPM files and find the proper version number. The following RPM file has a version of 1.30-426.0. This is the SDC RPM, but is only being used as the example of identifying the version number.

EMC-ScaleIO-sdc-1.30-426.0.el6.x86_64.rpm

Plug that version into the puppet/manifests/site.pp file. Important Note: If you are using the example site.pp files, make sure you update this version once you copy those files from puppet/manifests/examples to puppet/manifests/site.pp.

$version = '1.30-426.0'

Vagrant Deploy¶

Issue the following commands to retrieve the proper repository with necessary files and issue an ‘vagrant up’ command.

git clone https://github.com/emccode/vagrant-puppet-scaleio
cd vagrant-puppet-scaleio
mkdir -p puppet/modules/scaleio/files
cp RHEL6/EMC-ScaleIO-*.rpm puppet/modules/scaleio/files
cp ScaleIO_v1.30_GA/ScaleIO-GW\ \(IM+Rest-GW\)*.rpm puppet/modules/scaleio/files
vagrant up
vagrant status

Following this you should see the following Vagrant status report.

Current machine states:

puppetmaster              running (virtualbox)
tb                        running (virtualbox)
mdm1                      running (virtualbox)
mdm2                      running (virtualbox)

There are three network adapters that are auto-configured by default. Under the covers the 192.168.50.0/24 subnet is used for isolated ScaleIO communication.

Vagrant Customization¶

The Vagrantfile is responsible for creating the proper VMs, configuring them, and then running the respective Puppet commands to start the server or agents. The Vagrantfile and the ScaleIO Puppet module are completely separate and can be used independently.

Configure the Puppet Master details in the puppetmaster_nodes hash.

puppetmaster_nodes = {
  'puppetmaster' => {
    :ip => '192.168.50.9', :hostname => 'puppetmaster', :domain => 'scaleio.local', :memory => 512, :cpus => 1
  }
}

The scaleio_nodes hash holds the configuration of the ScaleIO nodes. Three nodes are needed at a minimum for ScaleIO to be configured. Here you can see that we have a Tie-Breaker and two MDMs. These nodes are all multi-role where they are also serving as servers and clients for data services. The ScaleIO Puppet module determines these component roles.

scaleio_nodes = {
  'tb' => { :ip => '192.168.50.11', :hostname => 'tb', :domain => 'scaleio.local', :memory => 1024, :cpus => 1 },
  'mdm1' => { :ip => '192.168.50.12', :hostname => 'mdm1', :domain => 'scaleio.local', :memory => 1024, :cpus => 1 },
  'mdm2' => { :ip => '192.168.50.13', :hostname => 'mdm2', :domain => 'scaleio.local', :memory => 1024, :cpus => 1 },
}

Vagrant Runtime¶

Vagrant status can be seen if you run ‘vagrant status’ from the directory with the Vagrantfile. Before modifying the Vagrantfile make sure that you destroy the existing ScaleIO VMs ‘vagrant destroy’.

At any point you can add nodes to the Vagrant file scaleio_nodes hash followed by a ‘vagrant up’ command. See Advanced Usage for more details.

Verify Node Deploy¶

You can verify at any time that the nodes have been configured using Puppet and ScaleIO commands.

From the node you can run a Puppet Agent command. The following command will run the agent again and verify all configuration is correct.

vagrant ssh node
sudo puppet agent -t

From an MDM node you can run ScaleIO commands. The following command will return general information.

vagrant ssh mdm2
sudo /bin/scli --login --username admin --password 'Scaleio123' --mdm_ip 192.168.50.12
sudo /bin/scli --query_all --mdm_ip 192.168.50.12

Use this command for a verbose listing of commands.

sudo /bin/scli --all --help

Puppet Node Redeploy¶

If you have configured a node prior and want to redeploy, make sure that you remove the certificate for the old node from the Puppet Master. Since ScaleIO requires that you install two MDM’s and a Tie Breaker as a base install, do not try and redeploy these nodes. This configuration must be done manually to get these nodes back online. Redeploys of other node types do not have the same ramifications.

vagrant ssh puppetmaster
sudo puppet cert clean node.scaleio.local

If the node still exists you also must remove the certificate from the node.

vagrant ssh node
sudo find / -name node.scaleio.local.cer -delete

Untar the RPM¶

The GUI requires that you download Java JRE to the system that is going to run the GUI. If you have brew installed you can leverage this to install via CLI.

brew update && brew cask install java

Verify installation with the following command.

which java

For Linux or OS X platforms, locate the GUI RPM named EMC-ScaleIO-gui-1.30-426.0.noarch.rpm. This RPM has a shell script and a JAR file. You can get the contents of the RPM without running an RPM command by using tar.

tar -zxvf EMC-ScaleIO-gui-1.30-426.0.noarch.rpm
./opt/emc/scaleio/gui/run.sh

Load the GUI¶

If Java is in the path, then the GUI should load. Enter admin for the username, and the password you configured in the site.pp file. The IP address is the primary MDM ip.

admin
Scaleio123
192.168.50.12

Add SDS Nodes¶

scaleio_nodes = {
  'tb' => { :ip => '192.168.50.11', :hostname => 'tb', :domain => 'scaleio.local', :memory => 1024, :cpus => 1 },
  'sds1' => { :ip => '192.168.50.14', :hostname => 'sds1', :domain => 'scaleio.local', :memory => 1024, :cpus => 1 },
  'sds2' => { :ip => '192.168.50.15', :hostname => 'sds2', :domain => 'scaleio.local', :memory => 1024, :cpus => 1 },
  'mdm1' => { :ip => '192.168.50.12', :hostname => 'mdm1', :domain => 'scaleio.local', :memory => 1024, :cpus => 1 },
  'mdm2' => { :ip => '192.168.50.13', :hostname => 'mdm2', :domain => 'scaleio.local', :memory => 1024, :cpus => 1 },
}

$sio_sds_device = {
  'tb.scaleio.local' => {
    'ip' => '192.168.50.11',
    'protection_domain' => 'protection_domain1',
    'devices' => {
      '/home/vagrant/sio_device1' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
    }
  },
  'mdm1.scaleio.local' => {
    'ip' => '192.168.50.12',
    'protection_domain' => 'protection_domain1',
    'devices' => {
      '/home/vagrant/sio_device1' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
    }
  },
  'mdm2.scaleio.local' => {
    'ip' => '192.168.50.13',
    'protection_domain' => 'protection_domain1',
    'devices' => {
      '/home/vagrant/sio_device1' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
    }
  },
  'sds1.scaleio.local' => {
    'ip' => '192.168.50.14',
    'protection_domain' => 'protection_domain1',
    'devices' => {
      '/home/vagrant/sio_device1' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
    }
  },
  'sds2.scaleio.local' => {
    'ip' => '192.168.50.15',
    'protection_domain' => 'protection_domain1',
    'devices' => {
      '/home/vagrant/sio_device1' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
    }
  },
}

node /sds/ {
  class {'scaleio::params':
        password => $password,
        version => $version,
        mdm_ip => $mdm_ip,
        sio_sds_device => $sio_sds_device,
        sds_ssd_env_flag => true,
        components => ['sds'],
  }
  include scaleio
}

vagrant ssh puppetmaster
sudo /etc/init.d/puppetmaster restart

vagrant destroy -f
vagrant up

vagrant ssh mdm2
sudo puppet agent -t

Add New Devices to SDS¶

You can leverage this module to add new devices local to the SDS nodes to be consumed by the SDS service. Consult ScaleIO documentation for the requirements for new storage to SDS devices. Here we will show how to add same sized devices to all configured SDSs.

The following snippet shows the devices key with a single device (default) and two optional configuration parameters, size and storage_pool.

'devices' => {
'/home/vagrant/sio_device1' => {
        'size' => '100GB',
        'storage_pool' => 'capacity'
},

Below is the full sio_sds_device hash with mutliple devices per SDS.

$sio_sds_device = {
  'tb.scaleio.local' => {
    'ip' => '192.168.50.11',
    'protection_domain' => 'protection_domain1',
    'devices' => {
      '/home/vagrant/sio_device1' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
      '/home/vagrant/sio_device2' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
    }
  },
  'mdm1.scaleio.local' => {
    'ip' => '192.168.50.12',
    'protection_domain' => 'protection_domain1',
    'devices' => {
      '/home/vagrant/sio_device1' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
      '/home/vagrant/sio_device2' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
    }
  },
  'mdm2.scaleio.local' => {
    'ip' => '192.168.50.13',
    'protection_domain' => 'protection_domain1',
    'devices' => {
      '/home/vagrant/sio_device1' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
      '/home/vagrant/sio_device2' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },

vagrant ssh puppetmaster
sudo /etc/init.d/puppetmaster restart

vagrant ssh node
sudo puppet agent -t

Notice: /Stage[main]/Scaleio::Login/Exec[Normal Login Class]/returns: executed successfully
Notice: mdm2.scaleio.local
Notice: /Stage[main]/Scaleio::Sds_first/Notify[mdm2.scaleio.local ]/message: defined 'message' as 'mdm2.scaleio.local '
Notice: mdm1.scaleio.local
Notice: /Stage[main]/Scaleio::Sds_first/Notify[mdm1.scaleio.local ]/message: defined 'message' as 'mdm1.scaleio.local '
Notice: tb.scaleio.local
Notice: /Stage[main]/Scaleio::Sds_first/Notify[tb.scaleio.local ]/message: defined 'message' as 'tb.scaleio.local '
Notice: /Stage[main]/Scaleio::Sds/Exec[Add SDS mdm1.scaleio.local device /home/vagrant/sio_device2]/returns: executed successfully
Notice: /Stage[main]/Scaleio::Sds/Exec[Add SDS mdm2.scaleio.local device /home/vagrant/sio_device2]/returns: executed successfully
Notice: /Stage[main]/Scaleio::Sds/Exec[Add SDS tb.scaleio.local device /home/vagrant/sio_device2]/returns: executed successfully
Notice: /Stage[main]/Scaleio::Sds_sleep/Exec[Add SDS mdm1.scaleio.local Sleep 30]/returns: executed successfully
Notice: Finished catalog run in 36.98 seconds

Add Volume to SDC¶

You can leverage this module to manage the creation and mapping of Volumes to SDCs. For this you there is a sio_sdc_volume hash in the site.pp file. Notice how we added a key for volume2 specifying the size_gb, protection_domain, storage_pool, and which client SDC’s to advertise the volume to.

$sio_sdc_volume = {
  'volume1' => {
    'size_gb' => 8,
    'protection_domain' => 'protection_domain1',
    'storage_pool' => 'capacity',
    'sdc_ip' => [
      '192.168.50.11',
      '192.168.50.12',
      '192.168.50.13',
    ]
  },
  'volume2' => {
    'size_gb' => 16,
    'protection_domain' => 'protection_domain1',
    'storage_pool' => 'capacity',
    'sdc_ip' => [
      '192.168.50.11',
      '192.168.50.12',
      '192.168.50.13',
    ]
  }
}

vagrant ssh puppetmaster
sudo /etc/init.d/puppetmaster restart

vagrant ssh mdm2
sudo puppet agent -t

Notice: /Stage[main]/Scaleio::Login/Exec[Normal Login Class]/returns: executed successfully
Notice: /Stage[main]/Scaleio::Volume/Exec[Add Volume volume2]/returns: executed successfully
Notice: /Stage[main]/Scaleio::Map_volume/Exec[Add Volume volume2 to SDC 192.168.50.12]/returns: executed successfully
Notice: /Stage[main]/Scaleio::Map_volume/Exec[Add Volume volume2 to SDC 192.168.50.11]/returns: executed successfully
Notice: /Stage[main]/Scaleio::Map_volume/Exec[Add Volume volume2 to SDC 192.168.50.13]/returns: executed successfully
Notice: Finished catalog run in 4.74 seconds

Global Parameters¶

A global parameter gets set in the following section. These parameters can also be set individually per node if desired. These variable names are used later on in the node configuration section.

$version = '1.30-426.0'
$mdm_fqdn = ['mdm1.scaleio.local','mdm2.scaleio.local']
$mdm_ip = ['192.168.50.12','192.168.50.13']
$tb_fqdn = 'tb.scaleio.local'
$tb_ip = '192.168.50.11'
$cluster_name = "cluster1"
$enable_cluster_mode = true
$password = 'Scaleio123'
$gw_password= 'Scaleio123'

SDS Devices¶

The SDS Devices represent the block storage devices on an SDS node. As part of the parameters you also specify which protection domain an SDS node takes part in. The devices themselves can be block devices (partitions) that have a pre-determined size or files on file systems that will be truncated to create the sparse space.

This is an example of a hash that represents a node.

'FQDN' => {
  'ip' => 'IP of NODE',
  'protection_domain' => 'Protection Domain',
  'devices' => {
    'Device Path1' => {  'size' => '100GB',
                                      'storage_pool' => 'Name of Storage Pool'
                                    },
    'Device Path2' => {  'size' => '100GB',
                                      'storage_pool' => 'Name of Storage Pool'
                                    },
  }

This are working hashes that represents nodes.

$sio_sds_device = {
  'tb.scaleio.local' => {
    'ip' => '192.168.50.11',
    'protection_domain' => 'protection_domain1',
    'devices' => {
      '/home/vagrant/sio_device1' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
    }
  },
  'mdm1.scaleio.local' => {
    'ip' => '192.168.50.12',
    'protection_domain' => 'protection_domain1',
    'devices' => {
      '/home/vagrant/sio_device1' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
    }
  },
  'mdm2.scaleio.local' => {
    'ip' => '192.168.50.13',
    'protection_domain' => 'protection_domain1',
    'devices' => {
      '/home/vagrant/sio_device1' => {  'size' => '100GB',
                                        'storage_pool' => 'capacity'
                                      },
    }
  },
}

SDC Volumes¶

The volumes section represents the actual consumable device that a client has available to carve space from. Here you specify the Protection Domain and Storage Pool. You can also specify any abritrary size in GB. The sdc_ip represents an array of SDC IPs that will have access to the volume.

$sio_sdc_volume = {
  'volume1' => { 'size_gb' => 8,
  'protection_domain' => 'protection_domain1',
  'storage_pool' => 'capacity',
  'sdc_ip' => [
      '192.168.50.11',
      '192.168.50.12',
      '192.168.50.13',
    ]
  }
}

Call-Home¶

The Call-Home section is used to configure the Call-Home service.

$callhome_cfg = {
  'email_to' => "emailto@address.com",
  'email_from' => "emailfrom@address.com",
  'username' => "monitor_username",
  'password' => "monitor_password",
  'customer' => "customer_name",
  'smtp_host' => "smtp_host",
  'smtp_port' => "smtp_port",
  'smtp_user' => "smtp_user",
  'smtp_password' => "smtp_password",
  'severity' => "error",
}

Node Configuration¶

This is the node configuration. With each node statement there is a regular expression that determines what configuration is applied to nodes with specific FQDNs. The required configuration parameters differ depending on the node type. Since the MDM and TB nodes specified here have mixed components there are more variables specified that needed for those node types.

See the Global Parameters section for details of other parameters listed below. The components parameter is the only one that specifies something unique per node.

Notice that the nodes are using a Regular Expression (/tb/) which matches any node that checks in with TB in the name. The nodes can be duplicated and completely customizable based on site naming preferences. Notice the components line where we specify (tb,sds,sdc,gw) meaning the node has is multi-role as a Tie-Breaker, SDS, SDC, and Gateway.

node /tb/ {

class {‘scaleio::params’:

password => $password, version => $version, mdm_ip => $mdm_ip, tb_ip => $tb_ip, callhome_cfg => $callhome_cfg, sio_sds_device => $sio_sds_device, sds_ssd_env_flag => true, components => [‘tb’,’sds’,’sdc’],

} include scaleio

}

node /mdm/ {

class {‘scaleio::params’:

password => $password, version => $version, mdm_ip => $mdm_ip, tb_ip => $tb_ip, cluster_name => $cluster_name, sio_sds_device => $sio_sds_device, sio_sdc_volume => $sio_sdc_volume, callhome_cfg => $callhome_cfg, components => [‘mdm’,’sds’,’sdc’,’callhome’],

} include scaleio

}

node /sds/ {

class {‘scaleio::params’:

password => $password, version => $version, mdm_ip => $mdm_ip, sio_sds_device => $sio_sds_device, sds_ssd_env_flag => true, components => [‘sds’],

} include scaleio

}

node /sdc/ {

class {‘scaleio::params’:

password => $password, version => $version, mdm_ip => $mdm_ip, components => [‘sdc’],

} include scaleio

}

node /gw/ {

class {‘scaleio::params’:

gw_password => $gw_password, version => $version, mdm_ip => $mdm_ip, components => [‘gw’],

} include scaleio

}

Requires¶

• Tested with Puppet 3.7.2
• Internet access to download RPMs
• ScaleIO 1.30-426+

To Be Done¶

• Test updates
• Masterless Puppet Agents
• Fault Sets
• Licensing

Special Settings¶

• Puppet 3.7+

• puppet.conf [main] - parser = “future”