Is jLite still maintained?

[tdoherty]

Hi Oleg,

Is Jlite still maintained? It has been quite a while since the last post.

If it is then I will send a further question.

Thanks.

Regards,

Tom.

[wj Vriend]

Same question from my side.

We are just in the process of replacing jlite with the glite
voms-proxy tools and making a custom python interface.

jlite does not support proxies with more then 512 bits, if I remember
correctly, and that will be a problem in the near future.

Best regards,

Willem-Jan Vriend.

> --
> Вы получили это сообщение, поскольку подписаны на группу "jlite".
> Чтобы отменить подписку на эту группу и больше не получать от нее сообщения,
> отправьте письмо на электронный адрес jlite+un...@googlegroups.com.
> Чтобы отправлять сообщения в эту группу, отправьте письмо на электронный
> адрес jl...@googlegroups.com.
> Чтобы зайти в группу, перейдите по ссылке
> http://groups.google.com/group/jlite.
> Чтобы настроить другие параметры, перейдите по ссылке
> https://groups.google.com/d/optout.

[Oleg Sukhoroslov]

Hi guys,

I haven't used jLite for my projects for a while, so it was not updated.
jLite clearly needs an update since both gLite and its Java APIs have
been evolved in the recent years.
I can invest some time to do this in the nearest future.
Let me know if you have any specific requests.

Regards,
Oleg

[tdoherty]

Hi Oleg,

Thanks for your speedy reply.

I have used jLite extensively and indeed have built a Grid submission\monitoring
framework around it. So thanks for an excellent API.

It is true that time is moving forward in the Grid middleware world which might
mean my time with using the frameworks I have built is limited. There is a shift
away from using WMS's in the GridPP (UK) HEP world - which is the infrastructure
my VO uses (as a non LHC Physics VO) and soon I may be one of the only users
of the WMS's that are available which would mean they would be decommissioned.
There also seems to be a move away from CREAM CE's to ARC CE's which I have
not tested my workflows on (I use parametric jdl's).

My immediate problem is that the UK eScience grid certificate authority has moved
to using SHA-2 signing.

This results in the following error:

jlite.GridAPIException: org.globus.io.urlcopy.UrlCopyException: UrlCopy transfer failed. [Caused by: Authentication failed [Caused by: Defective credential detected [Caused by: [JGLOBUS-99] CRL verification failed [Caused by: no such algorithm: SHA256withRSA for provider BC]]]]
        at jlite.GridSessionImpl.submitJob(GridSessionImpl.java:651)
        at jlite.GridSessionImpl.submitJob(GridSessionImpl.java:465)
        at soho.grid.job.submit.App.run(App.java:893)
        at soho.grid.job.submit.App.main(App.java:92)
Caused by: org.globus.io.urlcopy.UrlCopyException: UrlCopy transfer failed. [Caused by: Authentication failed [Caused by: Defective credential detected [Caused by: [JGLOBUS-99] CRL verification failed [Caused by: no such algorithm: SHA256withRSA for provider BC]]]]
        at org.globus.io.urlcopy.UrlCopy.copy(UrlCopy.java:522)
        at jlite.GridSessionImpl.uploadFilesToGridFTP(GridSessionImpl.java:717)
        at jlite.GridSessionImpl.submitJob(GridSessionImpl.java:576)

What I was trying to figure out was how old the Bouncy Castle jar that is bundled with
jLite is - because BC does seem to provide the SHA256withRSA algorithm at the
moment. I tried to upload the latest jar with the bundle but to no avail - I guess it
needs some internal configuration but I will attempt to configure it on my client code
if that is possible.

Thanks very much for any help you can give me.

Regards,

Tom.

[Oleg Sukhoroslov]

Hi Tom,

A couple of questions concerning your problem:

1. Which WMS server did you access when got this error?
Possibly I can try it out if it supports ESR VO.

2. Do you see this error only during uploading files via GridFTP?
If so, the job should be successfully registered by WMS client right before upload, which narrows down the problem to GridFTP client code.

Regards,
Oleg

[Oleg Sukhoroslov]

Willem-Jan,

You are right, old version of VOMS Java API used by jLite supports only 512 bits proxies.

I've just tested recent version of VOMS Java API, it generates 1024 bits proxies by default and supports configuration of key length.

So it can be solved by upgrading jLite with dependencies.

Regards,

Oleg

среда, 2 июля 2014 г., 19:16:10 UTC+4 пользователь wjvriend написал:

Same question from my side.

We are just in the process of replacing jlite with the glite
voms-proxy tools and making a custom python interface.

jlite does not support proxies with more then 512 bits, if I remember
correctly, and that will be a problem in the near future.

Best regards,

Willem-Jan Vriend.

2014-07-02 16:49 GMT+02:00 tdoherty <tomdoh...@googlemail.com>:
> Hi Oleg,
>
> Is Jlite still maintained? It has been quite a while since the last post.
>
> If it is then I will send a further question.
>
> Thanks.
>
> Regards,
>
> Tom.
>
> --
> Вы получили это сообщение, поскольку подписаны на группу "jlite".
> Чтобы отменить подписку на эту группу и больше не получать от нее сообщения,

> отправьте письмо на электронный адрес jlite+unsubscribe@googlegroups.com.