active python environment do not match the version of the binaries...not able to update to 3.5

[HaroldDoran2]

I recently upgraded ArcPro to 3.5 and the clone of the environment all show 3.5. I also have the latest API 2.4.1. My problem was that code I had written to upload to AGOL no longer works. The error message I get is:

The version of the binaries in the active Python environment do
not match the version of the binaries in the ArcGIS application.

Active environment version : 3.4
arcgispro-py3 environment version : 3.5

I've gone through all the suggested approaches in these forums to update the active environment to no avail. 

the suggested approach (https://pro.arcgis.com/en/pro-app/latest/arcpy/get-started/activate-an-environment.htm) did not work

I've tried using Conda (e.g., conda proup) with no luck either.

What am I missing here? Any thoughts would be greatly appreciated and if anyone knows how to roll back an ArcPro update I'd really like to know.

Thank you all

H

[HaydenWelch]

o you have the ability to re-upload the code to AGOL using the new environment? I'm not sure exactly how AGOL handles python code, but if it bundles the package data with the code you could probably get around it by just re-uploading using the new environment.

Rolling back isn't too difficult either, just go to https://my.esri.com and go to this tab:

I'd also make sure that the clone environment you are using is referencing the correct arcpy package:

[HaroldDoran2]

Thanks for your quick reply. Unfortunately, our government license doesn't allow my to link my enterprise esri account to the my esri so I cannot download the previous version and install it. I did make sure the cloned environment and all the components were using 3.5 yet it still thinks it is in 3.4.

[HaydenWelch]

That page is just old installer downloads, if you still have your old installer lying around you can just uninstall Arc and re-run that installer and you'll get a downgrade. I keep a folder of the installers so I can roll back if something is messed up.

[DavidSolari · Accepted solution]

I've never had luck upgrading Pro environments after a new version drops, I just delete all my custom environments and build them from scratch. Have you given that a try? You can run conda list to get a list of packages for all your environments so you don't forget to reinstall anything.

[HaroldDoran2]

Thanks David that is exactly what I had to do...wiped everything away and started over and it cleared up the problem i.e., got rid of the old instance and now points to the correct environment.

thanks