------------------------------------------------------------------
Steven C. Timm, Ph.D  (630) 840-8525
[EMAIL PROTECTED]  http://home.fnal.gov/~timm/
Fermilab Computing Division, Scientific Computing Facilities,
Grid Facilities Department, FermiGrid Services Group, Assistant Group Leader.

On Thu, 26 Jun 2008, Lengyel, Florian wrote:

Perhaps this is what has happened: from
http://www.globus.org/mail_archive/gt-user/2007/09/msg00154.html
Charles Bacon wrote the following reply to a similar query of mine:

"I'm sorry it took me so long to see what's happening. You're right, it was 
something that got setup when the GLOBUS_LOCATION was first installed. I had missed 
that the job executed correctly and that only the Cleanup directive was failing.

When you install GT, it creates $GLOBUS_LOCATION/etc/gram-service/ 
globus_gram_fs_map_config.xml. That file tells GRAM what GridFTP servers to use 
for file staging and cleanup operations. I assume that has the old hostname 
embedded in it. Changing the hostname in there and restarting the container 
should fix it.


Charles"


The file above already has the correct host name for me, and my error
is slightly different.

The globus toolkit setup might not pick up GLOBUS_HOSTNAME, assuming this is
set to the desired interface.

I suppose I wimped out in the end by using the VDT, but that's another story.

It should be noted here that my globus is coming from  the VDT as well.
I did set GLOBUS_HOSTNAME before the install, and got a lot farther
that way than I did previously.

Steve Timm



Florian Lengyel, Ph.D.
Assistant Director for Research Computing
Department of Information Technology
and Adjunct Professor, Ph.D. Program in Computer Science
The CUNY Graduate Center
365 Fifth Avenue, Room 4435  New York, NY 10016
Phone: (212) 817-7374               FAX: (212) 817-1614
WWW: http://research.gc.cuny.edu
Email: [EMAIL PROTECTED]



-----Original Message-----
From: [EMAIL PROTECTED] on behalf of Steven Timm
Sent: Thu 6/26/2008 12:32 AM
To: [email protected]
Subject: [gt-user] Configuring globus, multi-ip situation


I have the following

A host with two public IP's

fnpc3x1.fnal.gov  131.225.167.18
fnpcosg1.fnal.gov 131.225.166.2

"hostname" returns fnpc3x1.fnal.gov

I want to offer the globus web services only on the 2nd ip.
fnpcosg1.fnal.gov.

hostcert.pem/hostkey.pem have host cert with fnpcosg1.fnal.gov
in the subject.  so does containercert.pem/containerkey.pem
DNS is OK both ways.

Globus pre-ws works ok, so does gsiftp.

on a globusrun-ws I get the following:

globusrun-ws -submit -F fnpcosg1.fnal.gov:9443 -Ft Condor -J -s -c
/usr/bin/id
Delegating user credentials...Done.
Submitting job...Done.
Job ID: uuid:a0ee8f04-4338-11dd-8cba-001422086c92
Termination time: 06/27/2008 04:30 GMT
globusrun-ws:
globus_service_engine.c:globus_l_service_engine_session_started_callback:2744:
Session failed to start
globus_xio_gsi.c:globus_l_xio_gsi_read_token_cb:1335:
The peer authenticated as
/DC=org/DC=doegrids/OU=Services/CN=fnpcosg1.fnal.gov.Expected the peer to
authenticate as /CN=host/fnpc3x1.fnal.gov

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

the interesting thing is that the job in question *does* get submitted
to condor, and run, but I get no stdout/stderr back.

I have grepped every single file in globus and see no reference
to the base ip anywhere in any config file.  What if anything did
I do wrong?

Steve Timm


------------------------------------------------------------------
Steven C. Timm, Ph.D  (630) 840-8525
[EMAIL PROTECTED]  http://home.fnal.gov/~timm/
Fermilab Computing Division, Scientific Computing Facilities,
Grid Facilities Department, FermiGrid Services Group, Assistant Group Leader.




Reply via email to