Working with Variables

[knorman]

I’m working on my first WFA development project and have ran into a hurdle I can’t seem to jump over. I ‘m looking for the way to take a location a user specifies and turn it into a local cluster name.  From there I’ll be selecting an aggregate to create a new volume in for a new VMware datastore.

I created the short bit of code below to translate datacenter location to the correct site cluster name as a couple sites have multiple cluster but only one for VMware.  My issue is feeding this newly assigned cluster name into the rest of the workflow so it uses the $ClusterName variable specified below throughout the workflow.  I’m not necessarily looking for a complete solution but rather the general way to make this work.

param (

[parameter(Mandatory=$false, HelpMessage="Datacenter Location")]

  [string]$DCLocation,

[parameter(Mandatory=$false, HelpMessage="Name of Cluster")]

  [string]$ClusterName

)

if ($DCLocation = 'Valley Forge')

            {$ClusterName = 'Amazon'}

elseif ($DCLocation = 'Denver')

            {$ClusterName = 'Phoenix'}

elseif ($DCLocation = 'Ft. Worth')

            {$ClusterName = 'Griffin'}

[dhruvd]

Hi knorman,

If the version of WFA that you are using is 2.2, you can make use of the powershell commandlets Add-WfaWorkflowParameter and Get-WfaWorkflowParameter. These will help you to add a parameter in one of the commands, which can then be accessed in any of the commands that follow.

You will have add "ClusterName" as a parameter in the command where you are translating the DCLocation to ClusterName in the following manner:

$ParameterName = "ClusterName"

$ParameterValue = $ClusterName

Add-WfaWorkflowParameter -Name $ParameterName -Value $ParameterValue

You can then get this parameter in the other command codes in the following way:

$ParameterName = "ClusterName"

$val = Get-WfaWorkflowParameter -Name $ParameterName

Note: You will not be able to access $ClusterName in other places in the workflow such as filters/loops. This variable will be scoped at command level only.

[adaikkap]

Hi Norman,

What you are really trying to do is resource selection based on location of the user who invoke the workflow.

For this you you will somehow need do a mapping of the location to user to resource. In such situation we generally recommend to use playground database to do such persistent mapping.

• If the mapping is mostly static, then you can use a playground DB and store the mapping from location to cluster name
• Or, you could use a function to return cluster name given the location and then use that to load the “Cluster by key” in a no-op command
• After that the rest of the workflow can use the cluster variable where it would be properly filled i.e cluster.name

Other way is

• Store in playground DB the mapping
• First user input can be location $Location
• Once chosen, the next locked query based user input named $ClusterName will be based on $Location (This is query based, but will always have single item in drop-down)

Solution Courtesy, Shailaja

Regards

adai

[olson]

Greetings,

                There was a previous conversation that included something that would solve your problem with a function. With a function you can create this sort of hash map.

def regionMap(region) {

  return ['Americas' : 'am', 'EMEA' : 'em', 'APAC' : 'ap', 'Switzerland' : 'ch', 'Singapore' : 'sn', 'Hong Kong' : 'hk'].get(region);

}

def locationMap(location) {

return ['Singapore':'sg','Tokyo':'tk','Sydney':'sy','Hong Kong':'hk','New York':'ny','London':'ln','Pune':'pn','Mumbai':'mu','Jakarta':'jk','Seoul':'se','Bangkok':'bk','Shanghai':'sh','Kuala Lumpur':'kl','Beijing':'bj','Taipei':'tp','Bangalore':'bg','Atlanta':'at','Boston':'bs','Buenos Aires':'ba','Calgary':'cg','Cayman Islands':'ci','Chicago':'cc','Dallas':'da','Houston':'hs','Las Angeles':'la','Mexico':'mx','Miami':'mi','Philadelphia':'ph','Raleigh':'ra','San Francisco':'sf','Toronto':'to','West Palm Beach':'wp','Paris':'pa','Dubai':'du','Doha':'do','Dublin':'db','Poland':'po','Madrid':'md','Milan':'ml','Turkey':'tu'].get(location);

}

To call it in your workflow you can simply use

(String)(regionMap($region)

volumeName

if ( $workflowMethod == 'Specified' ) { return ($INPUT_Volume) } else { return 'v' + (String)(regionMap($region)) + (String)(businessUnitMap($bu)) + (String)(useCaseMap($useCase)) }

vfilerName

if ( $workflowMethod == 'Specified' ) { return ($INPUT_VfilerName) } else { return 'vf' + (String)(regionMap($region)) + (String)(businessUnitMap($bu)) + (String)(useCaseMap($useCase)) + (String)(environmentMap($env)) }

[trentino123]

In the workflow setup, I would add 2 CONSTANTS, 1 is the LOCATION and another is the CLUSTER. Both have a string separated by commas ( they are an array ).

LOCATION = "location1,location2,location3,location4"

CLUSTER = "cluster1,cluster4,cluster1,cluster1"

Then with a function we could get an index number of the location and place then place the cluster in a variable ( CLUSTER ) we could use to list the aggregates.

[trentino123]

If working with a CSV file is not a problem ( to have a new datasource you could use in your SQL queries within the user inputs ), here is a 5min video that explains how to do it :

http://community.netapp.com/t5/OnCommand-Storage-Management-Software-Articles-and-Resources/Video-Building-your-first-script-based-custom-datasource-u...