There are at least 3 different connectors we can use with Dynamics CE instances when creating a Flow:
- CDS Connector
- Dynamics 365 Connector
- Http Connector
They are not identical, though, and, from what I have gathered so far, we may need to keep a few things in mind when choosing between the 3:
1. When using a CDS connector, make sure your environment has been upgraded
Otherwise, you won’t be able to see your Dynamics CE database from that environment. For example, here is what showed up when I tried creating a flow in the old Default environment which was never upgraded:
And here is what shows up when I’m working with a flow in my Dynamics CE environment:
2. There are rumors that Microsoft has stopped working on the Dynamics 365 Connector
That said, most of the Dynamics 365 connector functionality seems to be available in the CDS connector. Although, there is no “Create or Update” trigger in the CDS connector, and it’s not unusual when we need the same processing to happen in response to either of those events.
If you are using a CDS connector, you may, then, consider starting to use nested flows:
Which would, basically, look like this:
And yes, personally I think that’s quite a complication for somebody used to the native Dynamics workflows. Which is beyond the point, though, since those are different “technologies”.
However, unlike Dynamics connector, CDS connector has “Current” environment option which is why it can be packaged into solutions. You move it to another environment, and it’s still “Current”, so it just works.
3. HttpConnector is a low-level approach, and, just like any other low-level approach, it’s more complicated to implement, but, of course, it gives you most of the control
Mostly, it seems, we’ll be dealing with these two problems:
- There is no seamless authentication – here is how you can try working around it: https://community.dynamics.com/crm/b/mscrmshop/archive/2018/02/18/calling-d365-apis-from-logic-apps
- Forming the requests and parsing the responses – both CDS and Dynamics connectors have a rather good idea of what the environment looks like (entities, types, etc). Http Connector, on the other hand, will make no assumptions at all, so you’ll need to know the entity names, you’ll need to know how to write Web API queries, etc.
But, in return, you’ll be able to do everything(?) that can possibly be done with Web API