Upgrade vs Update: deleting rows?

[ultrAslan]

Hello,  I need to update a solution (instead of upgrade) due to several reasons*. Since I have never used "update" (because until now to "upgrade" was pretty enough for deployment),

should I be aware of the database in Dataverse? Will "update" remove the rows? Or are the rows not affected by "update"?
Thank you!

[Amik]

@ultrAslan - entity/table rows are not impacted when upgrading or updating a solution. The Upgrade or Update action is not the same as deleting a Managed Solution.

An Upgrade means the existing Solution in the target environment is uninstalled and replaced with a new version. Any components which do not exist in the new solution are uninstalled from the solution you're replacing.

An Update still replaces the existing Solution in the target environment with the newer solution, but any components which do not exist in the new Solution will continue to exist.

Further reading: https://learn.microsoft.com/en-us/power-apps/maker/data-platform/update-solutions 

[ultrAslan]

I tried Update, too, but I am still facing the following error: Solution upgrade fails due to a previously pending upgrade - Microsoft Dataverse | Microsoft Learn

As the sandbox environment was deleted, we lost our root unmanaged solutions. To be still able to maintan and edit the published Apps (in managed solutions) in the Live/Production Environment, we have built new Solutions around the exported Applications (from Prod Environment to new Sandbox Environment). Because the solutions were set up as manaeged solutions, we were not able to export them. But however, exporting the Apps (which are included in a managed solution) is possible. So we export them from Prod Environment and import into the new sandbox environment and built up the exact same solution as it's in the production environment. For another solution/app, this has worked. But now, in this case, it stucks with the error:

"Solution manifest import: FAILURE: Holding solution <Solution name>_Upgrade exists. Cannot update solution."

In the documentation  ( Solution upgrade fails due to a previously pending upgrade - Microsoft Dataverse | Microsoft Learn ), I have read this:
"To work around the issue, it's necessary to complete the previously pending upgrade by resolving any errors. Or, you can delete the previous upgrade solution. However, this might lead to data loss if new data comes through tables and relationships in the upgrade. Once the pending upgrade is either completed or deleted, you can proceed with the new upgrade."

Can I delete the solution in the overview which has now the name "<SolutionName>_Upgrade" while the original "<Solutions Name>" (without _Upgrade) still exists
WITHOUT losing any data in dataverse? I mean, both of them are managed solutions. Losing data would be worst case scenario...

Thank you for any help

---

[Amik]

@ultrAslan - I would not recommend deleting a managed solution without taking a back up of the DB at minimum. In this situation I will always wait for the existing upgrade to complete, prior to executing anything else.

[ultrAslan]

Yes, I was afraid that would happen.

Unfortunately, the Upgrade_Layer has been stuck for 3 hours. If anyone knows a workaround to fix this situation without losing data, I am open to any help. Thanks again.

[ultrAslan]

In the documentation it says "However, this might lead to data loss if new data comes through tables and relationships in the upgrade. ".


I am not interested in any data which comes through the sandbox environment, in fact it's empty there. 
The table properties are the same as it was in the previous one. Would deleting "pending upgrade solution"  cause a loss of data in the already existing tables which are NOT new?

[Amik]

@ultrAslan - not something I have cared to find out.  I have added @Pstork1 who might be able to share some insight.

[ultrAslan]

Thank you 🙂

[Pstork1]

I'm not sure there is an easy way to recover from this if you deleted the original unmanaged solutions. I suspect the easiest way will be to do the following.

1) Export the managed solution from production as an unmanaged solution

2) Import that unmanaged solution into a new developer environment

3) Export the new developer solution as a managed solution

4) Remove the managed solution from production and re-install the new managed solution

Be sure to backup the data first, but unless the data source is Dataverse you shouldn't lose the tables or any data.