After a DMA update (internal build) an existing setup using Process Automation no longer works.
How do we start to debug? Where to look?
The setup uses a permanent booking with a specific process.
This process needs to start on-demand. To do this an Automation script exists to generate and throw a token.
All of this was working fine and was not changed.
In an info event I can see that the automation script to throw the token runs, but nothing happens after this.
Where do I start looking to get an idea of what is going wrong?
DMA version:
Prev: 10.1.2.0-9839-20210115-internal
New: 10.1.7.0-10523-20210616-release
I would start by looking at the queue element(s), specifically the one to which the token is sent.
- is the token listed in the incoming table?
- do you have entries in the registration table?
- do you have exceptions in the element log of the queue element?
The original PA 1.2.2 suffered from compilation errors in the queue elements with DM 10.1.7.0 (released version) and CU1 was created for that (RN30206). If you have still the same issues with the latest CU, I would advise creating an issue task for the compilation problems.
Thanks! I had PA 1.2.1 installed. I now updated to 1.2.2 CU2.
Most likely the the problem is resolved by the DMA restart that is triggered by the update of PA. We will see with the next DMA update if the root cause is gone with the latest PA version or if there is still something else going on.
Hi Mieke,
When a PA workflow is not working, we start debugging this issue by checking the following items:
- Are the queue elements correctly configured? Check if the parameter 'Resource Pool' is pointing to a valid resource pool. I remember that we had an issue that we lose this configuration after a DMA restart. This has been fixed since PA 1.2.1
- Is the Resource Pool correctly configured? Check that all the resource pools that will be used in your PA workflow contains the property 'ProcessAutomation' set to True. This requirement is mandatory since PA 1.2.2
- Are you using gateways? Can you double check that all the rules are correctly configured? Can you also double check that the node property 'PA_FunctionType' has been set to 'Gateway'?
- Can you try creating a booking and double check that all the nodes contains resources that can be selected?
- If the manual booking has been triggered correctly, can you check double on which queue element the tokens are getting stuck? When tokens are stuck, you will see them listed in the table 'Touking Routing'.
Log files:
- Booking log file: If you have already a permanent booking running, check the booking log file if there are issues related to the booking life cycle.
- Script log file (SLAutomation.txt): This log file will give you additional information if there is an issue in the profile load scripts or the SRM_LSO_ProcessAutomation script
- Queue element log files: These log files will give you additional information in case there are issue with the token handling
Jan-Klaas,
Looks like the upgrade broke something in building/compiling the code of the queue elements:
…|SLManagedScripting.exe|ManagedInterop|ERR|-1|44|No CodeBuilder found for cookie0
…|SLProtocol – 12904 – SRM_Queue_PA PLM.SLC PLM Book Resources|3648|CQAction::Run|ERR|-1|QAction [90] triggered by [pid=8/idx=-1/pk=/user=] failed. (0x80040221)
Input: new = <NULL>
Input: old = <NULL>
Input: extra = <NULL>