Activity Chaining Not Working the second time of upload a file (File Upload in a GridLayout)

...s/subprocess used so I expect it to work since every flow lines is set to activity chaining.
Also, the Enable activity-chaining into all initial nodes in the sub-process checkbox is selected on the Setup Tab.

Thanks in advance for your help on this.

There are time limits for a chain to complete as well as a max number of unattended nodes. Have a look here forum.appian.com/.../Configuring_Activity-Chaining.html for more information. Is the 2nd upload going through more nodes than the first one and breaking the 50 limit?

Hi Rod,
Its just the same number of nodes 26 total nodes(includes nodes from subprocess). I tried testing if subprocesses are the culprit. I removed the subprocess that parses contents from excel file and takeout the script task from it and placed to the main process and it seems the chaining from that instance did not broke. It's working. I just wonder since that subprocess has just a single script task, it should not take a long time, not longer than the limit and it should not break the activity chaining.
That single script task calls 8 expression rules. Each parses a particular data and save to output CDTs.

Now I return another subprocess that deletes the removed files from the UI. It gets back the same issue where the chaining breaks. This subprocess has 5 unattended nodes 3 script tasks, 1 delete document node(runs n times depending on no. of deleted documents) and 1 delete data from entity node.

If I take it out again and placed these 5 nodes to the main process it defeats the purpose of having subprocess. What could you advice to fix this?

Are there any other MNI nodes in the chain ? Each instance of the MNI will count as a chain. Is it possible that you're then hitting 50 with that on the second time round? Perhaps not clearing out a variable so it's doubling up

Hi Rod, I have this delete document node that can run on specific times depending on the number of documents to delete. I tried deleting a row (document cdt row) so it should just run 1 instance and could not count to reach 50. It seems if there's an MNI the chaining breaks. Should I set the Document cdt to null after an execution of MNI?

Hi, can you verify, by monitoring the failing instance, where's the flow stuck? You may have a problem in a XOR gateway with more than one input: it has different behaviours when reached for the second time in a flow, waiting for both inputs before validating conditions for output flows.
If this is the case you should be able to see the flow stuck at the culprit gateway.

You shouldn't need to, is the process returning to the form but not chaining? Or as @albertom32 has mentioned is it getting stuck somewhere

It says "Action Completed Successfully" after deleting. It does not automatically show the form (activity chaining not working) so I have to open it on the task to access the task form again. We noticed on the first try, we're able to delete the row and get back to the form automatically(chaining working). But on succeeding tries, activity chaining breaks and the Action Completed Successfully shows. I would like to ask if succeeding tries accumulates/counts the number of nodes executed? If Yes, can this be cleared or reset?

The count is between non assigned tasks. So returning to a user input form clears the count and you start again from 0

@rodgraham yes, that's what I'm expecting too. But based from what happened, it seems that it's counting the previous nodes executed even after returning it to the user input task. I attached the screenshot, In there, you can see the blue lines, which has been first executed (parse excel) then the green line is the 2nd execution(delete attachment), the red line represents that this node has been executed in both execution (parse excel, then delete attachment). As per my understanding, I counted the subprocesses for save data and generate memo(both synchronous and activity chained) plus the other nodes in deleting attachment, and it's not reaching 50 times...but what happened is, it's closing automatically and chaining seems to break. Below is the log I found:
--> 2017-01-20 00:24:26 [PX011] {pe420.kdb ,310655} (Default) WARN .a.p.PROCESS.i "[P268437469-ICMS Generate Memo] [Activity Chaining] Could not chain to attended activity. Limit of 50 unattended nodes executed was reached."