Mailing-List: contact issues-help@camel.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@camel.apache.org
Date: Tue, 4 Jul 2017 12:28:00 +0000 (UTC)
From: "Claus Ibsen (JIRA)" <jira@apache.org>
To: issues@camel.apache.org
Message-ID: <JIRA.13074573.1495645856000.166228.1499171280236@Atlassian.JIRA>
In-Reply-To: <JIRA.13074573.1495645856000@Atlassian.JIRA>
References: <JIRA.13074573.1495645856000@Atlassian.JIRA> <JIRA.13074573.1495645856493@jira-lw-us.apache.org>
Subject: [jira] [Commented] (CAMEL-11321) Can CamelContext startup faster
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
archived-at: Tue, 04 Jul 2017 12:28:10 -0000


    [ https://issues.apache.org/jira/browse/CAMEL-11321?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16073577#comment-16073577 ] 

Claus Ibsen commented on CAMEL-11321:
-------------------------------------

As Caffeine cache takes 150-200 millis to startup etc we can try to have some LRUCacheFactory which initializes this in the background while Camel startup. We can look at where we use LRUCache and limit this to only when really needed so there are maybe some parts during startup that dont need that and can just use a regular bounded Map or something.

> Can CamelContext startup faster
> -------------------------------
>
>                 Key: CAMEL-11321
>                 URL: https://issues.apache.org/jira/browse/CAMEL-11321
>             Project: Camel
>          Issue Type: Improvement
>          Components: camel-core, camel-spring-boot
>            Reporter: Claus Ibsen
>             Fix For: Future
>
>
> [~lb] have some thoughts on if we can make CamelContext startup faster, especially when using Spring Boot. Spring Boot itself is not so fast, but should be faster in 2.x.
> For example we could consider camel-core to not do any classpath scanning for components that are provided OOTB in camel-core, eg if using "log" endpoint, then only check the spring registry if any custom bean of that, and if not, then we know "log" is from camel-core and then we know its class name already and dont need to scan the classpath.
> We could take that one step further for the entire Camel release and have a plugin that generate/keep java source file up to date from camel-core, which has complete mapping of all component-name=component-class.
> We can also do some profiling and see if there is some hot-spots. There is also the revised work that Zoran does in starting Camel on spring / spring-boot.


--
This message was sent by Atlassian JIRA
(v6.4.14#64029)