Return-Path: X-Original-To: apmail-aries-dev-archive@www.apache.org Delivered-To: apmail-aries-dev-archive@www.apache.org Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by minotaur.apache.org (Postfix) with SMTP id D838B17C90 for ; Wed, 16 Sep 2015 09:52:46 +0000 (UTC) Received: (qmail 3641 invoked by uid 500); 16 Sep 2015 09:52:46 -0000 Delivered-To: apmail-aries-dev-archive@aries.apache.org Received: (qmail 3433 invoked by uid 500); 16 Sep 2015 09:52:46 -0000 Mailing-List: contact dev-help@aries.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: dev@aries.apache.org Delivered-To: mailing list dev@aries.apache.org Received: (qmail 3194 invoked by uid 99); 16 Sep 2015 09:52:45 -0000 Received: from arcas.apache.org (HELO arcas.apache.org) (140.211.11.28) by apache.org (qpsmtpd/0.29) with ESMTP; Wed, 16 Sep 2015 09:52:45 +0000 Date: Wed, 16 Sep 2015 09:52:45 +0000 (UTC) From: "Christian Schneider (JIRA)" To: dev@aries.apache.org Message-ID: In-Reply-To: References: Subject: [jira] [Created] (ARIES-1405) Make sure bundle goes into graceperiod if services needed for JPA are not present MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit X-JIRA-FingerPrint: 30527f35849b9dde25b450d4833f0394 Christian Schneider created ARIES-1405: ------------------------------------------ Summary: Make sure bundle goes into graceperiod if services needed for JPA are not present Key: ARIES-1405 URL: https://issues.apache.org/jira/browse/ARIES-1405 Project: Aries Issue Type: Improvement Components: JPA Affects Versions: jpa-2.1.0 Reporter: Christian Schneider Assignee: Christian Schneider Fix For: jpa-2.2.0 Currently Coordinator as well as EntityManager services are injected using a manually created service proxy. The problem with this is that we can not stop blueprint from starting up the user context this way. So the jpa enabled beans start even if their prerequisites are not yet present. This can lead to calls blocking at runtime until the service timeout. The error reporting is also bad as you only know something went wrong after the first call and the timeout. So the idea is to use correct blueprint metadata for the service refs. This makes blueprint go into graceperiod until the services are present. The diag command will then also show what is missing. The current approach with the Beanprocessor does not support this approach. Se we need to switch to a ComponentDefinitionRegistryProcessor as recommended by Guillaume Nodet. -- This message was sent by Atlassian JIRA (v6.3.4#6332)