Belaid,

Please also reply to gt-user, not just to me.
You probably missed a question i put inline, which would be good to know
for me:

The system is designed as follows:

1- A grid service is used to take the necessary information to run a job
2- The same grid service creates the scripts and the RSL file
3- The same grid service than creates a thread that:
  a) submits the job to a computational grid, and
  b) listens to the job notifications, using GramListener, from the
computational grid.
After several tries, b) is the part that causes the problem. If I do not
listen to the notifications (batch mode),
everything works file, but when I do, the container error shows up after
I run the service.

What containers have the problem: those who host your grid service,
or those who are targeted by your grid-service (and don't host your
grid service), or both, or what?

Martin


Belaid MOA wrote:
Indeed, I am using  the following classes:

org.globus.exec.client.GramJob;
org.globus.exec.client.GramJobListener;


With best regards.
Belaid Moa.
----------------------------------------
Date: Thu, 18 Sep 2008 16:13:35 -0500
From: [EMAIL PROTECTED]
To: [EMAIL PROTECTED]; [email protected]
Subject: Re: [gt-user] ERROR container.

Ok, i'm pretty sure i know that the problem is
http://bugzilla.globus.org/globus/show_bug.cgi?id=3849

You are the second person within a few days hitting this.
I'll have to talk to the core guys about this.

How do you submit the jobs in your grid service: Do you use GramJob?
What is GramListener? Is that something we provide?

Martin


Belaid MOA wrote:
Hi Martin,
The system is designed as follows: 1- A grid service is used to take the necessary information to run a job
2- The same grid service creates the scripts and the RSL file
3- The same grid service than creates a thread that:
  a) submits the job to a computational grid, and
b) listens to the job notifications, using GramListener, from the computational grid. After several tries, b) is the part that causes the problem. If I do not listen to the notifications (batch mode), everything works file, but when I do, the container error shows up after I run the service.
What containers have the problem: those who host your grid service,
or those who are targeted by your grid-service (and don't host your
grid service), or both, or what?


The grid service, as well as others, is accessed through a web portal.
The jobs are submitted to a completely different machine.
The grid services work all the time but the command line tools such globus-credential-delegate, globusrun-ws, etc
do not work and show the gridmap error after the grid service runs.
Thank you very much Martin for your help. With best regards.
Belaid Moa.
Date: Wed, 17 Sep 2008 19:00:33 -0500
From: [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Subject: Re: [gt-user] ERROR container.

Hang on, what do you mean by "in a separate thread"?
Do you run that service in the same container and you submit jobs in that
service using the Gram Java API GramJob and you let GramJob listen for
notifications?
And you get the problems once you submitted a single job in that service?

Thanks, Martin

Belaid MOA wrote:
Hi Martin,
Thanks for looking after the problem. Once I restart the container
everything works fine including the job submission
command lines (globusrun-ws, globus-job-submit, etc.). But once I
run my
intensive grid service that uses GramListener to listen to job status
notification in a separate thread, I start getting the error in the
container logfile, and command line job submission does not work
anymore. My grid services still work from the portals. I can still
submit the jobs through my grid service. The error I get after running
my service is:

------------------------------------------------------------------------------------------------
[EMAIL PROTECTED] testData]$ globusrun-ws -submit -host -S -F
mars.csc.uvic.ca -c /bin/uname -a
Submitting job...Failed.
globusrun-ws: Error submitting job
globus_soap_message_module: SOAP Fault
Fault code: soapenv:Server.userException
Fault string:

org.globus.wsrf.impl.security.authorization.exceptions.AuthorizationException:
Policy decision failed [Caused by: No gridmap file]

-------------------------------------------------------------------------------------------------------
Any one trying to submit to my machine will get the same error.

With best regards.
Belaid Moa.




Date: Wed, 17 Sep 2008 15:09:02 -0500
From: [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
CC: [EMAIL PROTECTED]; [email protected]
Subject: Re: [gt-user] ERROR container.

Belaid,

so everything works fine until this error shows up in the container
logfile,
and after that ALL submission requests fail with the gridmap error
message?
Martin

Belaid MOA wrote:
Thanks a lot Charles for your quick reply. The services work,
however,
when I run the command
line: globusrun-ws -submit -S -F mars.csc.uvic.ca -c /bin/uname -a
Submitting job...Failed.
globusrun-ws: Error submitting job
globus_soap_message_module: SOAP Fault
Fault code: soapenv:Server.userException
Fault string:

org.globus.wsrf.impl.security.authorization.exceptions.AuthorizationException:
Policy decision failed [Caused by: No gridmap file]

Although I have a gridmap file
(/etc/grid-security/grid-mapfile) with
the right permissions(644), I keep getting the error above.
This only
happened after the ERROR Container.GSIServiceThread showed up!

With best regards.
Belaid Moa.


CC: [email protected]
From: [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Subject: Re: [gt-user] ERROR container.GSIServiceThread
Date: Tue, 16 Sep 2008 14:06:06 -0500

I'm pretty sure you can make this happen just by telnetting
to the
container port and disconnecting. In other words, if anyone's
ever
doing scans of your open ports, you'll get this exception when it
connects to your server. The server tries to read from the
network
socket, but gets an EOF because the scanner never says anything
on the
connection.

Are you experiencing any problems with your services, or just
curious
about this logfile entry?


Charles

On Sep 16, 2008, at 1:38 PM, Belaid MOA wrote:

Dear All,
For some strange reasons my container (GT 4.0.6) keeps
showing the
following error from time to time. I noticed that this was
reported
as a bug
(http://bugzilla.globus.org/bugzilla/show_bug.cgi?id=3449)
but I do not know whether it was fixed or not. Any help on this
issue is very appreciated.

=
=
=
=

======================================================================
2008-09-16 11:24:35,739 ERROR container.GSIServiceThread
[ServiceThread-11,process:147] Error processing request
java.io.EOFException
at
org
.globus
.gsi
.gssapi
.net

.impl.GSIGssInputStream.readHandshakeToken(GSIGssInputStream.java:56)
at
org

.globus.gsi.gssapi.net.impl.GSIGssSocket.readToken(GSIGssSocket.java:
60)
at
org

.globus.gsi.gssapi.net.GssSocket.authenticateServer(GssSocket.java:
122)
at

org.globus.gsi.gssapi.net.GssSocket.startHandshake(GssSocket.java:142)
at

org.globus.gsi.gssapi.net.GssSocket.getOutputStream(GssSocket.java:
161)
at
org
.globus

.wsrf.container.GSIServiceThread.process(GSIServiceThread.java:99)
at

org.globus.wsrf.container.ServiceThread.run(ServiceThread.java:291)
=
=
=
=

======================================================================

With best regards.
Belaid Moa.

_________________________________________________________________


------------------------------------------------------------------------


------------------------------------------------------------------------
Get your information fix on your phone. With MSN Mobile you get
regular
news, sports and finance updates. Try it today!

------------------------------------------------------------------------
Upgrade to Hotmail Plus and share more photos with bigger attachments. Click here to find out how Click here to find out how

_________________________________________________________________



Reply via email to