Trigger Flow within a FLow/dynamically.

You can pass the variables.
look a little bit above where the trigger information is, and you will see the tab 'Process Variables'.

Try putting a variable in the Service field on the WorkObject.

Hi Terry - I'm trying to follow the suggestions you were given to solve your issue, based on the print screen you provided you want a variable in the "Process:" drop down list on the Trigger Node?  Is that correct?  

If so, I'm assuming since there is no 'orange light bulb' you can't use a variable.  The drop down list contains the processes that are available in the rich client.

I would suggest using two trigger nodes with a branch node before them using 'your criteria' to determine which trigger node go to.

However Williams & Henley know a heck of a lot more then I do, so there is probably a better way to do this using a variable.  I'm learning too, so please let me know what works for you.

Thanks

Joan

Define the variable you want to pass (e.g. VAR1) in the start node in both Process A and Process B.
Then in the trigger node in Process B, click on the 'Process Variables' tab in the properties pane.
You should see the variable VAR1 listed at the bottom, with a checkbox next to it to pass it.
When Process A hits the trigger node for Process B, it will pass VAR1's value to the start node in Process B.

Oh... you're talking about passing a variable's value from one process flow to another process flow. That's not how I interpreted her question, but I understand your answer now.  Thank you.

Joan

Actually I think I misinterpreted the question

David was actually more on target.
Depending on how dynamic it needs to be and whether or not you want to do maintenance within the flow (bad) or via service definition (good), there are two ways to do it:

1. Flow: use branching logic with multiple branches to multiple trigger nodes using ProcessSync/ProcessAsync linked to specific process(es). This of course means that you have to know what to expect in the flow/variables, etc. and anytime you want to add a new trigger you have to maintain and re-deploy the flow, etc.

2. Service (as David suggested): You would have to set up a service and instead of ProcessSync/ProcessAsync you use ServiceSync/ServiceAsync and hook the trigger up to a server, and pass the variable as criteria. Then in the service definition you would map/route to the appropriate flow using the criteria.

Hi all,  thanks for the replies, I should of been more explicit on my description on what I needed,  but in the end that is what I needed.  I need to dynamically submit a Process Flow.   I was hoping there was a way to do that with out using the branch node.  So far I do not need to pass param to the named flows.  

What I need to solve, is to allow for auto scheduling of our process flows.  The scheduling of the Process Flows is not quite dynamic enough for the what we need.  

My thought was to create a SQL defined user table with the process flows that need to run at the required times.  Use a process flow to read that sql defined table and submit the required flow using the trigger node.  I did not want to use the branch node as it could get pretty involved and a possible maintenance nightmare.    

From your replies, it should work, and I just need to find out how many process flows we are talking about.  

Thanks for information, it is appreciated.

Terry

 

 

Hi Terry - I'm curious to know what type of auto scheduling you need to do that IPA can't handle?

Joan

Hi Joan,   this is going to be hard to explain, I will try to give a general high level overview.  We are very new to Lawson and moving our payroll and benefits from a different platform.  On our current platform, we are totally auto scheduled, there is no user intervention with our job scheduling system, except for job set up and if there is maintenance to be done.  We have quite a few job dependencies between jobs.  IE.  Job A cannot run with Job B, If Job A fails, Job B cannot run until Job A is fixed.   But Job C can run because it has no dependencies on Job A or Job B.    Also,  it gets complicated if one of our production sites is running behind and cannot get their data to us for payroll run.   We can put jobs on hold, which will automatically adjust the times on the schedule.   

This is very very high level.  But from what we have seen, and talking to people that use Lawson, most jobs are manually submitted and run by the the payroll people.  

Also, because we are fully automated on our other platform, we have to incorporate the Lawson jobs to be run as part of the auto-scheduling process.   

Hope this explains this a bit. 

Thanks

Terry

  

 

Terry,

Just to throw my two cents in, you could set up something similar to what you describe with ProcessFlow (IPA, Integrator, etc.). You have to start somewhere, so we'll start with Job A. In Job A's process flow you could either write a single line to a text file, something like "PASS" or "FAIL". After Job A finishes, call up the process flow for Job B. Job B process flow runs and reads the file. If it's "PASS", continue on with Job B process flow. If "FAIL", have it send you a notification of the fail condition so that you could trouble shoot the issue. You could also have the other Jobs (Job C, D, E, etc.) read different files to get the conditions of other process flow runs. You could pass the "PASS"/"FAIL" as variables between flows, but this way you could manually start the process anywhere in between and have the rest of the Jobs run after that. In the scheduler piece you also have the grace periods that you could use if the flows didn't run at the scheduled time. If you were really desperate, you could put in timer loop conditions with Javascript into the flow to check for certain conditions (like if "PASS" or "FAIL" exists in the file). I don't recommend this as once you go down that path it makes it difficult for troubleshooting and increases the chances that flows would "hang".

Terry- there are many ways that this can be achieved using IPA.
Thinking out loud,
To what jamesraceson said above; instead of using JavaScript timer loops, you can now use the 'Wait' node functionality in IPA. It frees up the resources/ threads back to the pool and can tap back into the loop.
or you may use the File Channels/ BCI components to run a trail of interdependent flows or a combination of custom DB triggers and multiple flows doing this. Please feel free to contact me for more information.
And Welcome to the Infor/ Lawson Family!

Hello,  Thanks for all of the information, I will be looking at the information that you all have provided and dig into it and see if I can make it work for us.  The biggest issue we are going to have is that our job scheduling process runs under a zVSE system, and it does does not have a real nice way to communicate to SQL and flat files on other platforms or Lawson IPA nodes directly.   

Thanks for the input!

Terry

 

Thank you Terry for the explanation.  Good Luck to you ;-)

Hello, I know I would be readdressing this after a while.   I used the first suggestion just to get by for now, but need to get suggestion #2 to work.   Using dynamic variables. 

"David was actually more on target. 
Depending on how dynamic it needs to be and whether or not you want to do maintenance within the flow (bad) or via service definition (good), there are two ways to do it: 

1. Flow: use branching logic with multiple branches to multiple trigger nodes using ProcessSync/ProcessAsync linked to specific process(es). This of course means that you have to know what to expect in the flow/variables, etc. and anytime you want to add a new trigger you have to maintain and re-deploy the flow, etc. 

2. Service (as David suggested): You would have to set up a service and instead of ProcessSync/ProcessAsync you use ServiceSync/ServiceAsync and hook the trigger up to a server, and pass the variable as criteria. Then in the service definition you would map/route to the appropriate flow using the criteria."

I set up a service and using the trigger node, I do not know how to pass the variable as criteria. Then in the service definition you would map/route to the appropriate flow using the criteria. 

How can I send 'IPAtest1" to the service to run that flow?

Any help would be appreciated.

Thanks

Terry

 

You need to do this via the service configuration in rich client. Set up the service's "criteria name 1". Then assign each unique flow to the service based on the value of Criteria 1.
For example, attach:
unique flow IPATest1 to criteria value 1 = IPATest1
unique flow IPATest2 to criteria value 1 = IPATest2

In other words, you need to use the service combined with the criteria value to trigger the desired flow.
It's similar to a scenario where you might have different approval flows based on a requesting location.

Let me know if that makes sense or not...

Thanks John,  I did not get a chance to look at/try the info you provided here yet.