Launching, as a batch, new BPM processes

[USDA]

Is it possible for BPM 7.6 to batch create several new and separate BPM processes?  Can this be done by reading and incorporating data from an external (non BPM table) into a new BPM process/form, one record at a time?  Each new record data from the non BPM table will be used to create a new and separate BPM process, complete with new Folder IDs.  Thanks.

tim. . . .

[BMellert]

Do you mean processes -- or folders, even if in multiple processes?

Presuming you mean folders, even if in multiple processes, one method may be to raise flags.  The same flag name can be used in multiple processes if appropriate (be very sure/careful you aren't launching too much if using a common flag), or have unique flags for each process.

Then it would be a matter of reading the source and launching the appropriate flags.

[USDA]

It would be processes.  For example, an external table that is not connected to any BPM processes, call it Table A, has the same type of data fields (but no Folder Ids) as would be contained in a BPM process, call it BPM process (ABC).  Each record/line in Table A is read and then a separate and new BPM process is kicked off that has the unique Folder Id for each record that is batch read from Table A.  The result is that the ABC process will have added records added to its table that is in Table A.  Is this possible? 

[S Cutura]

Hi,

yes that is possible, no problem. You can build something like a scheduler runnining in defined intervalls which reads the records in Table A. For each record you can then raise a flag to initiate a new process folder of ABC. You can pass data (like an id for the record in Table A) within the flag call. Once a new process folder has been created you can then use the Id to "import" all required data from Table A into the new folder.

You can pass all the data in the flag call but I would recommend that when it is more than a couple of values to pull the data from Table A than pushing it into the process ABC.

[USDA]

Is the 'schedular', that you refered to, a Metastorm function?

[USDA]

Do you have any process examples that I can view?

[Jerome҉]

Have a look here:

http://metastorm.processmapping.com.au/post/Importing-Legacy-Data-1581765

http://metastorm.processmapping.com.au/post?id=3686994

[S Cutura]

Hi USDA,

The "Scheduler" is not a built in MBPM functionality. It is basically a small process which has just the purpose to run timed actions.

1. Create a new project

2. On your process map create a stage with a timed action

3. Determine the list of new folders to be created for each timed run (All records with the status e.g. "New")

4. Set the items you are processing to a status e.g. "In Progress"

5. For each record in your list run the raiseflag method

6. Set the status of the items in your table e.g. to "Completed".

I hope that helps.

Sascha

[BMellert]

Alternatively, a Windows Task scheduler could be used to raise flags.  I know some folks don't like that approach with some valid reason.  However, I've found the timed action approach results in "time creep" over time, at least using ## time after last updated.  (e.g. if something is intended to run at 0:00, using a timed action eventually results in it "creeping" to later times as time goes on.  One weekly midnight job is now about 20 minutes after.

I suspect there are other ways to keep it at a fixed time, but the last updated isn't necessarily the answer if a fixed time every time is required.

[USDA]

I looke at the links, but Form1 and Form2 were emply.