Thanks a lot. I hope to have a look later this week (🤞 )

Thanks again for this PR. I think it makes sense in this way, even though I am not completely happy with the syntax of readOnlyTxManager....(...., readOnly=true). It feels somehow like a duplication. On the other hand I see no way how the write-mode transaction manager in the executable query should know that it is about to join an ongoing transaction.

Something that is missing in the PR is the support for the reactive template manager (you're missing all the fun ;) )
Unfortunately there is no validate method on the reactive transaction manager that will help us (you) to verify that this is working, or I cannot find it right now.

Hi @meistermeier ,

Will add the reactive part. Thanks for reviewing 👍

Hi @meistermeier ,

As of now, AbstractReactiveTransactionManager does not support transactional validation as does AbstractPlatformTransactionManager.setValidateExistingTransaction.

So it does not make sense to add it now.

Sorry to come back to this that late. Had a longer look at this today and I am happy with this change.
Could you please rebase/squash the commits and add the Sign-Off? (https://spring.io/blog/2025/01/06/hello-dco-goodbye-cla-simplifying-contributions-to-spring)

No problem. Will do 👍

There you go @meistermeier.

Merged this PR to 8.0.x with 0bedafb and main (upcoming 7.5.x) with 5aa41e7
Thanks again