cayenne-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ari Maniatis (Jira)" <j...@apache.org>
Subject [jira] [Commented] (CAY-2638) Cayennne sends commit
Date Tue, 05 Nov 2019 05:49:00 GMT

    [ https://issues.apache.org/jira/browse/CAY-2638?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16967256#comment-16967256
] 

Ari Maniatis commented on CAY-2638:
-----------------------------------

Well, I learnt something here today too!

> Cayennne sends commit
> ---------------------
>
>                 Key: CAY-2638
>                 URL: https://issues.apache.org/jira/browse/CAY-2638
>             Project: Cayenne
>          Issue Type: Bug
>          Components: Core Library
>    Affects Versions: 4.1.RC2
>            Reporter: Tomas Jura
>            Priority: Major
>         Attachments: commit_log.txt
>
>
> The cayenne sends the commits to the database ( after a lazy fetch, ... ). The transaction
boundaries are defined by a superior business logic and ORM must not corrupt this logic and
boundaries. Imagine for example that the Cayenne is part of an XA transaction, where the commit
is controled by a transaction manager and the commit guided by two phase commit protocol.
I'm afraid that any attempt to control transaction boundaries from ORM sooner or later clash
with required boundaries.
> This design feature, automaticaly reject using cayenne as an ORM for any bit complex
project which want to use savepoint/rollback/commit commands. As far as I remember Cayenne
3 did not touched the transaction state. Why Cayenne 4 touching the transaction?



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

Mime
View raw message