camel-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Claus Ibsen <>
Subject Re: [DISCUSS] - Add camel-test-spring component
Date Wed, 25 Jan 2012 05:51:43 GMT
On Wed, Jan 25, 2012 at 12:08 AM, Christian Müller
<> wrote:
> I like this improvement. This is the right way, IMO.

Yeah let me create a JIRA ticket so we capture this.

> I also share the opinion from Hadrian about the CamelSpringTestSupport.
> It's great for our users (I'm also one of them :-) ) to test their routes.
> And we need of course a few tests to make sure it works. We may need a few
> more tests which have more an example character to demonstrate the XML DSL
> for our users. But in general, most of our tests should "only" use the
> CamelTestSupport.

Yeah me to as well. Most components should be fine using the plain
However a few could be using CamelSpringTestSupport in the nature of
being examples.
Many people ask how to do X with Camel and either want a Java based or
XML based example.

Also when people contribute components to Apache Camel, then they may
write unit tests
which are more Spring XML favored than plain Java based. So we cannot
control this.

However by removing the Spring pieces from camel-test, we may be able
to have a better overview
of which components that uses CamelSpringTestSupport, as they would
not compile anymore :)

> As already mentioned, after I resolved the issues I'm working on, I will
> have a look at our unit tests.
> Best,
> Christian
> On Tue, Jan 24, 2012 at 10:07 PM, Hadrian Zbarcea <>wrote:
>> Revision [3] is just the tip of the iceberg. There were other before
>> mostly related to using dynamic ports and proper cleanup that allowed the
>> forkMode='once'. One of the other components that would greatly benefit
>> from that is camel-jms.
>> On the hanging test on Windows, that's obviously an issue that has to be
>> resolved and has little to do with the improvements I mentioned. My only
>> Windows system is an Atom based netbook which I rarely use, I cannot
>> confirm or deny experiencing that. Christian and Willem are usually pretty
>> fast at resolving Windows issues.
>> Speaking of CI and jenkins, I don't know of any asf project that takes
>> that long to test and sucks so many resources. Fwiw, that's one of the
>> improvements on my agenda for 3.0.
>> Cheers,
>> Hadrian
>> On 01/24/2012 11:03 AM, Babak Vahdat wrote:
>>> Hadrian,
>>> you mentioned a really good point regarding the improvements by camel-cxf
>>> testing that I intended to bring into the discussion for today anyway:
>>> Last night I was looking at my monitor as the full build on Windows (build
>>> 216) was running and was really stumped as the test:
>>> org.apache.camel.component.**cxf.cxfbean.CxfBeanTest
>>> Was running, running and running again [1] for exactly 365 minutes, that's
>>> more than 6 hours!
>>> At the last Jenkins killed the build process and now I assume because of
>>> that (and maybe some other reasons as well) we're banned from the build on
>>> Windows due to previous failure(s) [2].
>>> I assume your mentioned revisions like [3] by your post.
>>> [1]****
>>> 216/console<>
>>> [2]****
>>> 217/console<>
>>> [3]**view=revision&revision=1151313<>
>>> Babak
>>> --
>>> View this message in context: http://camel.465427.n5.nabble.**
>>> com/DISCUSS-Add-camel-test-**spring-component-**tp5281207p5294755.html<>
>>> Sent from the Camel Development mailing list archive at
>> --
>> Hadrian Zbarcea
>> Principal Software Architect
>> Talend, Inc

Claus Ibsen
Twitter: davsclaus, fusenews
Author of Camel in Action:

View raw message