Hi folks,
On my return from Empower 2025, I have been experimenting with the DataAPI and have found a problem. I have three different windows services pushing events in via the Data API and two of them work just fine but the other is seeing the following message returned from the Data API:
Production protocol version for ** **** ***** does not exist
We see the same error if we delete a connector which has been generated by the DataAPI, although the connector in this example has never existed.
So two things:
- Why is it throwing this error?
- Is there a way of clearing the DataAPI's knowledge of what connectors it has built so we can start again?
Hi Chris,
If your services continued running overnight, it could be that a subsequent request created the connector, otherwise, I would check if it was some kind of synchronization issue. Are you testing on a single agent or on a cluster environment?
Hi Chris,
Could you tell us a bit more about how you send the requests to DataAPI?
Are all 3 services sending to a single connector? Or does each service send to a separate one?
My questions arise because we have encountered some edge cases where other error messages from similar validations would be returned.
These were typically associated with multiple concurrent requests, and with DataMiner being slower than usual, leading to timeouts and failures.
Regarding your second question, the best way of clearing things in DataAPI is to
- Stop the pushing of events to DataAPI
This will prevent DataAPI from recreating connectors as you are deleting them in the next steps - Delete the connectors from DataMiner via the Protocols and Templates module
- Restart DataAPI DxM service
This will clear any in-memory data - Resume the event pushing
At this point, DataAPI will behave as if it had never known the deleted connectors.
I came into work this morning and logged on only to find that the protocols had appeared overnight. How strange.
Any thoughts on why this might be and how to avoid it in the future?