Ivan,

Thanks for picking up this ticket!
I know that Michael has commented on your PR at
https://github.com/apache/fineract/pull/642 but I'll try to see if others
in community can also review.

Regarding your comments, you are correct that we had to replace Hibernate
and at the time OpenJPA was the best option but now EclipseLink has become
a more attractive option. As you probably saw, for Fineract CN we migrated
over to EclipseLink from Hibernate. Despite one of the key decision factors
for that migration was the better performance between EclipseLink and
PostgreSQL which doesn't factor in for Fineract 1.x, I do think we should
seriously consider migrating to EclipseLink as the benefits to EclipseLink
and the downsides of OpenJPA with the activity of its community waning are
significant.

Could a member of the community try to estimate level of effort for
migration? It would be good to explore pros/cons in terms of better
performance, long term support, easier compatibility versus the short term
disruption it would cause implementers.

Ivan, are you interested in doing a spike to see how much effort migration
would take? I'll dig up old webinar which outlined all the steps taken for
the hibernate to OpenJPA migration.

Ed

On Fri, Oct 11, 2019 at 5:56 AM Ivan Bondarenko <[email protected]> wrote:

> Hello!
>
> I have raw implementation for (https://issues.apache.org/jira/browse/FINER
> <https://issues.apache.org/jira/browse/FINERACT-724>FINERACT-724
> <https://issues.apache.org/jira/browse/FINERACT-724>FINERACT-724
> <https://issues.apache.org/jira/browse/FINERACT-724>ACT-724
> <https://issues.apache.org/jira/browse/FINERACT-724> Upgrade Spring Boot,
> Spring and Spring Security to their latest stable version).
> This version compiles and all integration tests are passing.
> But I have done quite horrible things for Spring5/OpenJpa integration.
> OpenJpa is a pain, but as I understand we can't use Hibernate due to
> license restrictions.
> But why we can't use Eclipse link?
> Also I would like propose to move from JAX-RS to Spring MVC.
>
> Could you have a look?
>
> Thanks,
> Ivan
>
>
> ср, 2 окт. 2019 г. в 10:08, Michael Vorburger <[email protected]>:
>
>> On Wed, Oct 2, 2019 at 8:07 AM Ivan Bondarenko <[email protected]>
>> wrote:
>>
>>> Hello Michael !
>>>
>>> Thank you, it was gradle daemon issue.
>>> I use Intellig Idea and this IDE works with gradle daemon by default.
>>>
>>
>> cool, glad we were able to solve that.
>>
>> I have one more question:
>>>
>>> Is there any reason for old version of tomcat in dev dependencies?
>>>
>>
>> not really - you can certainly try to upgrade it - pull requests (which
>> still pass ITs on Travis..) are certainly welcome. However, In terms of
>> focusing energy, I personally would much rather welcome any contributions
>> re. below to help us move off external WAR and Tomact and to full Spring
>> Boot (with embedded Tomcat), if you know what I mean. Moving to a newer
>> Tomcat version, but staying with the current model, and
>> gradle-tomcat-plugin, is of more limited value, IMHO.
>>
>>
>>> def tomcatVersion = '7.0.54'
>>> while in prod
>>> def tomcatVersion = '7.0.94'
>>>
>>> It is really interesting for me to help you with gradle-tomcat-plugin.
>>> How can I join to you in this work?
>>>
>>
>> read up on and start contributing (as in, send us Pull Requests..), or if
>> needed ask more specific technical questions re. these open issues, in
>> order of priority:
>>
>> 1. https://issues.apache.org/jira/browse/FINERACT-764 Run Integration
>> Tests using Spring Boot IT support instead of on Tomcat started separately
>> by gradle-tomcat-plugin
>>
>> 2. https://issues.apache.org/jira/browse/FINERACT-730 Fix failing gradle
>> task bootRun & can't run simple Spring Boot java -jar (without Tomcat)
>>
>> 3. https://issues.apache.org/jira/browse/FINER
>> <https://issues.apache.org/jira/browse/FINERACT-724>FINERACT-724
>> <https://issues.apache.org/jira/browse/FINERACT-724>FINERACT-724
>> <https://issues.apache.org/jira/browse/FINERACT-724>ACT-724
>> <https://issues.apache.org/jira/browse/FINERACT-724> Upgrade Spring
>> Boot, Spring and Spring Security to their latest stable version
>>
>> We are looking to working with you!
>>
>>
>>> Best regards,
>>> Ivan
>>>
>>>
>>>
>>>
>>> вт, 1 окт. 2019 г. в 19:08, Michael Vorburger <[email protected]>:
>>>
>>>> On Tue, 1 Oct 2019, 17:58 Ivan Bondarenko, <[email protected]> wrote:
>>>>
>>>>> Hello!
>>>>>
>>>>> I'm trying to run integration tests locally.
>>>>> And first time it fails on awaitSpringBootActuatorHealthyUp()
>>>>> "/fineract-provider/health" returns 404.
>>>>>
>>>>
>>>> You should check the log to see why..
>>>>
>>>> Next times tomcat fail to start with error:
>>>>> *Caused by: java.lang.IllegalStateException: Illegal class loader
>>>>> binding*
>>>>>
>>>>> (Full stack in attachment)
>>>>>
>>>>> I have searched for this problem and it looks like the issue in
>>>>> embedded tomcat plugin.
>>>>> (https://github.com/bmuschko/gradle-tomcat-plugin/issues/46)
>>>>>
>>>>
>>>> This is strange - two separate./gradlew runs should be completely
>>>> isolated.. the "next time" shouldn't be any different than the "first
>>>> time". Are you using Gradle's daemon? Don't- it doesn't work for us, with
>>>> the gradle-tomcat-plugin... Note that we disable it via
>>>> https://github.com/apache/fineract/blob/develop/gradle.properties -
>>>> how are you getting around that?? ;-)
>>>>
>>>> Could you advice me how to fix this issue?
>>>>> Should I use `env=dev` for integrationTests?
>>>>>
>>>>
>>>> Nope, not required.
>>>>
>>>> BTW we'd like to get rid of the gradle-tomcat-plugin all together. I've
>>>> partial WIP re that; details available upon request. Would you have any
>>>> interest in helping this project with that?
>>>>
>>>> Thanks,
>>>>> Ivan
>>>>>
>>>>

Reply via email to