It would be nice if Flow would support on-premise Active Directory like it supports already on-premise SQL Server.
I'm sure this would help multiple organizations to automate their HR and IT routine tasks (new employees, leaving employees, etc.).
I think I saw somewhere that you can install an on premise service, that would do that. and it wouldn't be hard, considering you can connect to any HTTPs endpoint. as long as you have the correct credentials.
Sure, we can install on-premise services and yes, I believe we can connect to any HTTPs endpoint as well. Still I'm not sure how to solve this without on-premise Active Directory support in Flow.
I see a number of vendors that market these onPremise connectors as preconfigured and out of the box for their own products. If you know how to do it, then you too can make money out of it. But as for microsoft helping you 🙂 good luck. lol.
Why don't you do what most companies do, and a one way trust your onPrem AD to Azure AD. and use E1 and E5 accounts licensed via AAD. that way your users are all in AAD automatically. but still normal updates happen in onPrem AD.
I think Azure AD does not synchronize users created in Azure AD to on-premise AD? I think syncronization of new users works only from on-premise AD to Azure AD.
Im in for this!
there is a need not only get info about users, but also about some other types of objects. Printers is a good example. A connector from microsoft would be very helpful, particularly that the gateway exist. The one could be extended easily.
a connector would be nice indeed: in the meantime perhaps running an azure runbook (starting an on prem powershell)
@ToniRantanen wrote:Hi! It would be nice if Flow would support on-premise Active Directory like it supports already on-premise SQL Server. I'm sure this would help multiple organizations to automate their HR and IT routine tasks (new employees, leaving employees, etc.). Best regards,Toni Rantanen
Did you manage to get any workaround to this in the end?
@domtuck96, unfortunately no. Perhaps there would have been a workaround for this but all workarounds sounded too technical or too much workload for me to complete so we are still doing this manually. 😞