We are seeing different formats for our ProcessDefinitionIDs between processes - sometimes just a GUID 69b50396-6256-11e8-9121-0242ac11001c and sometimes a concatenation of name:version:GUID CA-instruction-process:2:ca360213-618d-11e8-b470-0242ac11001b. We also see different versioning after deployment - with the concatenated id, the deployment number gets incremented in certain cases.
Deploying via REST to the same Camunda instance using the same deployment script, so I’m assuming there’s something different in the BPMN file configiration, but I can’t see it. The files are too big and too different to meaningfully post hear, but does anyone know what to look for or able to point me to documentation which may help? I haven’t found any mention of this in the online docs so far.
We have found the cause of this difference - if the BPMN process id attribute is too long (mine was 32 chars, I’m not sure what the limit is) it seems the engine reverts to just allocating a GUID as the process definition id; if it has space, it concatenates other related/useful info.
This is a bit opaque and it’s not clear to me whether there is any other impact, I assume not. It did cause us some in a development environment we are building for embedded forms.
I would always treat the definitionId as a random (unique) number. You can use a ProcessDefinitionQuery to find a process definition in the runtime database.