Latest 2GP Packaging Engine throws error with Apex Rollup: "RollupFullRecalcProcessor : Managed type that is released without a supertype can not subsequently extend another type"

[cdevelin]

Hi James,

I realise this appears to be a bug in the latest version of Salesforce's 2GP packaging / validation process, but overnight our inclusion of APEX Rollup into our package has failed with this error:

RollupFullRecalcProcessor : Managed type that is released without a supertype can not subsequently extend another type

Does this error make any sense to you? Before we log a ticket with SF support and try to get them to even grasp the problem, would be good to get your take on the likely source of the validation failure, and whether it is perhaps legitimate and something we should be doing differently when incorporating APEX rollup as part of our package, even though it has been packaging fine into each new version without error for months.

[jamessimone]

Hey @cdevelin, I would definitely file a case for this. That sounds like a breaking change that will need to be investigated post-haste. I haven't changed anything about RollupFullRecalcProcessor in ages; certainly not something that should be breaking packaging for you. Let me know the case number once you've submitted and I'll try to follow up internally as well

[cdevelin]

Thanks James, case number is 470764963

[cdevelin]

@jamessimone SF support have not made any progress on this error arising from the update to their packaging engine (I understand it was overhauled to support the new upgrade path from 1GP to 2GP), and it appears some rules were not the same between 1GP and 2GP.

I was wondering whether you encounter the same packaging error if you try generating your own updated Managed Package version of APEX Rollup?

[jamessimone]

There isn't a managed package version of Apex Rollup, but I will try tomorrow - if I can - to generate a beta package version for the next release to see if there are any issues.

[cdevelin]

Hi @jamessimone This Supertype issue is still unresolved in Salesforce’s packaging process.

We tried breaking ancestry and creating a fresh package which includes Apex Rollup, but has no ancestors; packaged fine with no errors.

We then tried to create an update to that fresh package which had no changes to Apex Rollup. Packaging fails with that supertype issue with Apex Rollup's full recalc processor. Seems to be a bug in the ancestry checking logic of Salesforce's packaging process around that particular design 😣

[jamessimone]

Thanks for the update. I am closing this issue - there's nothing I can do on my end to resolve it, though I do believe (and hope for your success) that continuing to follow up with support to press for a backwards-compatible fix to packaging

[sirephil]

@cdevelin we have just started to hit the same error when trying to create a new version of our 2GP managed package. Did you get any feedback from Salesforce Support on your case?

[cdevelin]

@cdevelin we have just started to hit the same error when trying to create a new version of our 2GP managed package. Did you get any feedback from Salesforce Support on your case?

@sirephil You will need to log a ticket with SF support, you can reference our current ticket if you like (#471000158)

• its a regression bug in the packaging engine
• the impact of the packaging bug is it will have omitted this APEX class reference (a global class which extends another global class) from your ancestry registry maintained internally by SF’s packaging validation system. There is no way for you to fix or overcome or workaround this yourself - SF need to manually fix this for you, and you can never generate an update package again until they do
• SF are working on a fix for the bug, but that is to stop it omitting this type of APEX class from the ancestry registry in future, the bug fix will not fix your ancestry registry, hence just waiting for the software fix wont correct your “corrupt” ancestry registry, you must log a ticket

[sirephil]

@cdevelin many thanks for that. Case creation on its way!

[jamessimone]

Thanks @cdevelin! I know this has been a long tail on closing this issue, but it sounds like thanks in part to your attention to it, this regression will soon be fixed.

[sirephil]

For anyone else impacted by this, in addition to cdevelin's case (#471000158) you can also cite our Salesforce Support case (#471105149).

[rajashekharrugi]

@cdevelin @sirephil @jamessimone

We seem to be hitting the same problem. Logged a case and under investigation. Any inputs please how to handle this sooner?