Skip Menu | Logged in as guest | Logout
 
Ticket metadata
Id: 3258
Status: resolved
Priority: 3/0
Queue: vdt-support

Fixed in: (no value)
Fix scheduled: NR

Owner: Scot Kronenfeld
Requestors: wbetts@bnl.gov
Cc:
AdminCc:

More about wbetts@bnl.gov
Comments about this user:
No comment entered about this user
This user's 10 highest priority tickets:
Groups this user belongs to:
  • Everyone

New reminder:

Created: Thu Jan 24 11:42:51 2008
Starts: Not set
Started: Not set
Last Contact: Wed Jan 30 17:02:54 2008
Due: Not set
Closed: Fri Feb 01 11:37:50 2008
Updated: Fri Feb 01 11:37:50 2008 by kronenfe



History Brief headersFull headers
Subject: RSV problem at STAR-BNL OSG site
Date: Thu, 24 Jan 2008 12:38:22 -0500
To: VDT support tickets via RT <vdt-support@OPENSCIENCEGRID.ORG>
From: Wayne Betts <wbetts@bnl.gov>
Download (untitled) / with headers
text/plain 553b
Hi Scott et al.,

I went through the RSV installation instructions from the OSG CE
installation guide, but something seems to have gone amiss.
Condor-devel got installed and is running (I think), but there is
probably some sort of environment conflict going on with the existing
condor installation on the gatekeeper (v. 6.8.4). The RSV jobs are
winding up in that condor queue, rather than the condor-devel queue, if
I'm interpreting things correctly.

Can we work on figuring out what went wrong and/or how to fix it?

Thank you,

Wayne Betts
Download (untitled) / with headers
text/plain 413b
Hi Wayne--

Thanks for following up our discussion today at the VDT Office Hours
with a ticket. I've assigned it to Scot Kronenfeld. He made the RSV
probes work with Condor, so this is right up his alley.

-alain

-----------------------------------------------------------------
Alain Roy vdt-support@opensciencegrid.org
VDT Support http://vdt.cs.wisc.edu/support.html
CC: Arvind Gopu <agopu@indiana.edu>
Subject: Re: [vdt-support #3258] RSV problem at STAR-BNL OSG site
Date: Thu, 24 Jan 2008 13:07:47 -0600
To: vdt-support@OPENSCIENCEGRID.ORG
From: Scot Kronenfeld <kronenfe@cs.wisc.edu>
Hi Wayne,
The reason I mentioned making sure that you are running condor_q on the
right condor is because it does not matter which condor_q binary you run (
i.e. the pre-installed one, or the one in $VDT_LOCATION/condor-devel/bin).
The important thing is what the env variable CONDOR_CONFIG is set to.

If you do the following:

cd $VDT_LOCATION
. setup.sh
echo $CONDOR_CONFIG

you should be pointing at your pre-installed Condor's config file.
Condor-Devel does not put anything into the global setup.sh file in order to
minimize conflicts with your main Condor install. At this point, if you run
condor_q, you should not see any RSV jobs. If you do, we'll have to resolve
that issue.

In order to see your Condor-Devel queue, do the following
. vdt/etc/condor-devel-env.sh
echo $CONDOR_CONFIG (this should now be pointing at condor-devel's config
file)
condor_q

Ideally, you would now see the RSV jobs (and only the RSV jobs).

Let me know if the RSV jobs are running in the wrong Condor instance, and if
so, we'll debug further.

Thanks,
Scot
Subject: Re: [vdt-support #3258] RSV problem at STAR-BNL OSG site
Date: Thu, 24 Jan 2008 14:47:21 -0500
To: vdt-support@OPENSCIENCEGRID.ORG
From: Wayne Betts <wbetts@bnl.gov>
Download (untitled) / with headers
text/plain 4.9k
Here is the state of things:

[root@stargrid02 OSG]# echo $VDT_LOCATION
/opt/OSG-0.8.0
[root@stargrid02 OSG]# echo $CONDOR_CONFIG
/etc/condor/condor_config <--- This is the pre-existing condor config
file, as expected
[root@stargrid02 OSG]# condor_q


-- Submitter: stargrid02.rcf.bnl.gov : <130.199.6.168:20971> :
stargrid02.rcf.bnl.gov
ID OWNER SUBMITTED RUN_TIME ST PRI SIZE CMD
142781.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142782.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142783.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142784.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142785.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142786.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142787.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142788.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142789.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142790.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142791.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142792.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142793.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142794.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142795.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
probe_wrapper.pl /
142796.0 wbetts 1/22 17:18 0+00:00:00 I 0 9.8
gratia-script-cons
142797.0 wbetts 1/22 17:18 0+00:00:00 I 0 9.8 html-consumer
142798.0 wbetts 1/22 17:18 0+00:00:00 I 0 9.8
rotate_html_files.

18 jobs; 3 idle, 0 running, 15 held


Now I look at the condor-devel queue:

[root@stargrid02 OSG]# . vdt/etc/condor-devel-env.sh
[root@stargrid02 OSG]# echo $CONDOR_CONFIG
/opt/OSG-0.8.0/condor-devel/etc/condor_config
[root@stargrid02 OSG]# condor_q


-- Submitter: stargrid02.rcf.bnl.gov : <130.199.6.168:36559> :
stargrid02.rcf.bnl.gov
ID OWNER SUBMITTED RUN_TIME ST PRI SIZE CMD

0 jobs; 0 idle, 0 running, 0 held
[root@stargrid02 OSG]#


This is unchanged since Tuesday when I tried to get RSV working. My
first guess is that condor-cron spawns a process to submit the jobs,
but the environment gets corrupted because the OSG setup scripts have
links to them in /etc/profile.d. Is that possible? I'll try to get a
crash course in condor-cron...

-Wayne


Scot Kronenfeld via RT wrote:
> Hi Wayne,
> The reason I mentioned making sure that you are running condor_q on the
> right condor is because it does not matter which condor_q binary you run (
> i.e. the pre-installed one, or the one in $VDT_LOCATION/condor-devel/bin).
> The important thing is what the env variable CONDOR_CONFIG is set to.
>
> If you do the following:
>
> cd $VDT_LOCATION
> . setup.sh
> echo $CONDOR_CONFIG
>
> you should be pointing at your pre-installed Condor's config file.
> Condor-Devel does not put anything into the global setup.sh file in order to
> minimize conflicts with your main Condor install. At this point, if you run
> condor_q, you should not see any RSV jobs. If you do, we'll have to resolve
> that issue.
>
> In order to see your Condor-Devel queue, do the following
> . vdt/etc/condor-devel-env.sh
> echo $CONDOR_CONFIG (this should now be pointing at condor-devel's config
> file)
> condor_q
>
> Ideally, you would now see the RSV jobs (and only the RSV jobs).
>
> Let me know if the RSV jobs are running in the wrong Condor instance, and if
> so, we'll debug further.
>
> Thanks,
> Scot
>
>
>
>
> ------------------------------------------------------------------------
>
> Hi Wayne,
> The reason I mentioned making sure that you are running condor_q on
> the right condor is because it does not matter which condor_q binary
> you run (i.e. the pre-installed one, or the one in
> $VDT_LOCATION/condor-devel/bin). The important thing is what the env
> variable CONDOR_CONFIG is set to.
>
> If you do the following:
>
> cd $VDT_LOCATION
> . setup.sh
> echo $CONDOR_CONFIG
>
> you should be pointing at your pre-installed Condor's config file.
> Condor-Devel does not put anything into the global setup.sh file in
> order to minimize conflicts with your main Condor install. At this
> point, if you run condor_q, you should not see any RSV jobs. If you
> do, we'll have to resolve that issue.
>
> In order to see your Condor-Devel queue, do the following
> . vdt/etc/condor-devel-env.sh
> echo $CONDOR_CONFIG (this should now be pointing at condor-devel's
> config file)
> condor_q
>
> Ideally, you would now see the RSV jobs (and only the RSV jobs).
>
> Let me know if the RSV jobs are running in the wrong Condor instance,
> and if so, we'll debug further.
>
> Thanks,
> Scot
Subject: Re: [vdt-support #3258] RSV problem at STAR-BNL OSG site
Date: Thu, 24 Jan 2008 14:24:50 -0600
To: vdt-support@OPENSCIENCEGRID.ORG
From: Scot Kronenfeld <kronenfe@cs.wisc.edu>
Download (untitled) / with headers
text/plain 6.6k
Wayne,
The osg-rsv startup script (at $VDT_LOCATION/post-install/osg-rsv) sources
$VDT_LOCATION/vdt/etc/condor-devel-env.sh. This sets CONDOR_CONFIG to point
at the condor-devel instance. Then it submits the jobs via condor_submit as
the user you run rsv as (which looks to be wbetts).

Since I can see from your last email that source
condor-devel-env.shsupplies the right value for CONDOR_CONFIG, one
possibility is that there
was an error when sourcing that file (permissions or something?). It does
not seem likely, but I am confused about how else the jobs could end up in
your main Condor's queue.

Have you tried stopping and starting osg-rsv since you noticed this error?

vdt-control --off osg-rsv
vdt-control --on osg-rsv

If the problem persists, you can try to add a debugging line in the osg-rsv
startup script ($VDT_LOCATION/post-install/osg-rsv). Right underneath where
condor-devel-env.sh is sourced, try printing $CONDOR_CONFIG:

##
## Source the Condor-Devel setup file
##
. $VDT_LOCATION/vdt/etc/condor-devel-env.sh
echo $CONDOR_CONFIG

Then try to stop and start osg-rsv again. You'll have to look at the bottom
of $VDT_LOCATION/vdt-install.log for the output of that echo command. Let
me know how that goes, and we'll go from there.

Thanks,
Scot


On Jan 24, 2008 1:51 PM, wbetts@bnl.gov via RT <
vdt-support@opensciencegrid.org> wrote:

> Here is the state of things:
>
> [root@stargrid02 OSG]# echo $VDT_LOCATION
> /opt/OSG-0.8.0
> [root@stargrid02 OSG]# echo $CONDOR_CONFIG
> /etc/condor/condor_config <--- This is the pre-existing condor config
> file, as expected
> [root@stargrid02 OSG]# condor_q
>
>
> -- Submitter: stargrid02.rcf.bnl.gov : <130.199.6.168:20971> :
> stargrid02.rcf.bnl.gov
> ID OWNER SUBMITTED RUN_TIME ST PRI SIZE CMD
> 142781.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142782.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142783.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142784.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142785.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142786.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142787.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142788.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142789.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142790.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142791.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142792.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142793.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142794.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142795.0 wbetts 1/22 17:18 0+00:00:00 H 0 9.8
> probe_wrapper.pl /
> 142796.0 wbetts 1/22 17:18 0+00:00:00 I 0 9.8
> gratia-script-cons
> 142797.0 wbetts 1/22 17:18 0+00:00:00 I 0 9.8 html-consumer
> 142798.0 wbetts 1/22 17:18 0+00:00:00 I 0 9.8
> rotate_html_files.
>
> 18 jobs; 3 idle, 0 running, 15 held
>
>
> Now I look at the condor-devel queue:
>
> [root@stargrid02 OSG]# . vdt/etc/condor-devel-env.sh
> [root@stargrid02 OSG]# echo $CONDOR_CONFIG
> /opt/OSG-0.8.0/condor-devel/etc/condor_config
> [root@stargrid02 OSG]# condor_q
>
>
> -- Submitter: stargrid02.rcf.bnl.gov : <130.199.6.168:36559> :
> stargrid02.rcf.bnl.gov
> ID OWNER SUBMITTED RUN_TIME ST PRI SIZE CMD
>
> 0 jobs; 0 idle, 0 running, 0 held
> [root@stargrid02 OSG]#
>
>
> This is unchanged since Tuesday when I tried to get RSV working. My
> first guess is that condor-cron spawns a process to submit the jobs,
> but the environment gets corrupted because the OSG setup scripts have
> links to them in /etc/profile.d. Is that possible? I'll try to get a
> crash course in condor-cron...
>
> -Wayne
>
>
> Scot Kronenfeld via RT wrote:
> > Hi Wayne,
> > The reason I mentioned making sure that you are running condor_q on the
> > right condor is because it does not matter which condor_q binary you run
> (
> > i.e. the pre-installed one, or the one in
> $VDT_LOCATION/condor-devel/bin).
> > The important thing is what the env variable CONDOR_CONFIG is set to.
> >
> > If you do the following:
> >
> > cd $VDT_LOCATION
> > . setup.sh
> > echo $CONDOR_CONFIG
> >
> > you should be pointing at your pre-installed Condor's config file.
> > Condor-Devel does not put anything into the global setup.sh file in
> order to
> > minimize conflicts with your main Condor install. At this point, if you
> run
> > condor_q, you should not see any RSV jobs. If you do, we'll have to
> resolve
> > that issue.
> >
> > In order to see your Condor-Devel queue, do the following
> > . vdt/etc/condor-devel-env.sh
> > echo $CONDOR_CONFIG (this should now be pointing at condor-devel's
> config
> > file)
> > condor_q
> >
> > Ideally, you would now see the RSV jobs (and only the RSV jobs).
> >
> > Let me know if the RSV jobs are running in the wrong Condor instance,
> and if
> > so, we'll debug further.
> >
> > Thanks,
> > Scot
> >
> >
> >
> >
> > ------------------------------------------------------------------------
> >
> > Hi Wayne,
> > The reason I mentioned making sure that you are running condor_q on
> > the right condor is because it does not matter which condor_q binary
> > you run (i.e. the pre-installed one, or the one in
> > $VDT_LOCATION/condor-devel/bin). The important thing is what the env
> > variable CONDOR_CONFIG is set to.
> >
> > If you do the following:
> >
> > cd $VDT_LOCATION
> > . setup.sh
> > echo $CONDOR_CONFIG
> >
> > you should be pointing at your pre-installed Condor's config file.
> > Condor-Devel does not put anything into the global setup.sh file in
> > order to minimize conflicts with your main Condor install. At this
> > point, if you run condor_q, you should not see any RSV jobs. If you
> > do, we'll have to resolve that issue.
> >
> > In order to see your Condor-Devel queue, do the following
> > . vdt/etc/condor-devel-env.sh
> > echo $CONDOR_CONFIG (this should now be pointing at condor-devel's
> > config file)
> > condor_q
> >
> > Ideally, you would now see the RSV jobs (and only the RSV jobs).
> >
> > Let me know if the RSV jobs are running in the wrong Condor instance,
> > and if so, we'll debug further.
> >
> > Thanks,
> > Scot
>
>
> --
> View ticket at <
> http://vdt.cs.wisc.edu/rt/Ticket/Display.html?user=guest&pass=guest&id=3258
> >
> VDT Support, vdt-support@ivdgl.org
>
Subject: Re: [vdt-support #3258] RSV problem at STAR-BNL OSG site
Date: Fri, 25 Jan 2008 16:26:53 -0500
To: vdt-support@OPENSCIENCEGRID.ORG
From: Wayne Betts <wbetts@bnl.gov>
Download (untitled) / with headers
text/plain 16.3k

Message body not shown because it is too large or is not plain text.

Subject: Re: [vdt-support #3258] RSV problem at STAR-BNL OSG site
Date: Fri, 25 Jan 2008 17:10:20 -0600
To: vdt-support@OPENSCIENCEGRID.ORG
From: Scot Kronenfeld <kronenfe@cs.wisc.edu>
Download (untitled) / with headers
text/plain 973b
>
>
> On a vaguely related note: the "Condor guy" here at BNL (Alex Withers)
> has suggested installing the latest stable Condor 7.x, replacing our
> current condor 6.8.4. Might that handle the RSV jobs and eliminate the
> need for condor-devel?
>
> -Wayne
>

Yes, this should work. We will be working on the RSV infrastructure for the
1.8.2 release to work with pre-installed Condor 7.0 instances, and mostly
eliminate the need for Condor-Devel.

For right now, with VDT 1.8.1, you should be able to set
VDTSETUP_CONDOR_DEVEL_LOCATION to the location of your Condor 7.0 install
(in this case, also being your main Condor install), and RSV will use this
location. I'm going to run a test for this type of install locally to make
sure there are not additional steps, and I'll get back to you about the
verdict. In the meantime, if you decide not to take this approach, let me
know and I will help you debug this issue regarding the 6.8.x install
further.

Thanks,
Scot
Subject: Re: [vdt-support #3258] RSV problem at STAR-BNL OSG site
Date: Mon, 28 Jan 2008 09:34:15 -0600
To: vdt-support@OPENSCIENCEGRID.ORG
From: Scot Kronenfeld <kronenfe@cs.wisc.edu>
Download (untitled) / with headers
text/plain 1.4k
On Jan 25, 2008 5:10 PM, Scot Kronenfeld <kronenfe@cs.wisc.edu> wrote:

>
> > On a vaguely related note: the "Condor guy" here at BNL (Alex Withers)
> > has suggested installing the latest stable Condor 7.x, replacing our
> > current condor 6.8.4. Might that handle the RSV jobs and eliminate the
> > need for condor-devel?
> >
> > -Wayne
> >
>
> Yes, this should work. We will be working on the RSV infrastructure for
> the 1.8.2 release to work with pre-installed Condor 7.0 instances, and
> mostly eliminate the need for Condor-Devel.
>
> For right now, with VDT 1.8.1, you should be able to set
> VDTSETUP_CONDOR_DEVEL_LOCATION to the location of your Condor 7.0 install
> (in this case, also being your main Condor install), and RSV will use this
> location. I'm going to run a test for this type of install locally to make
> sure there are not additional steps, and I'll get back to you about the
> verdict. In the meantime, if you decide not to take this approach, let me
> know and I will help you debug this issue regarding the 6.8.x install
> further.
>

Hi Wayne

I ran a VDT 1.8.1 test locally using a pre-existing Condor 7 installation
for both Condor and Condor-Devel. The RSV jobs and some manual jobs I
submitted all ran fine.

During the install, you usually set VDTSETUP_CONDOR_LOCATION to point at
your pre-existing install. When installing OSG-RSV, you will also need to
set VDTSETUP_CONDOR_DEVEL_LOCATION to point at this same install.

Let me know if you have any questions.

Thanks,
Scot
Subject: Re: [vdt-support #3258] RSV problem at STAR-BNL OSG site
Date: Wed, 30 Jan 2008 16:05:33 -0500
To: vdt-support@OPENSCIENCEGRID.ORG
From: Wayne Betts <wbetts@bnl.gov>
Download (untitled) / with headers
text/plain 1.3k
Hi Scot,

After upgrading condor to 7.0 and configuring it to allow 'local'
universe jobs, the RSV jobs appear to be running successfully. Can you
check to see if the Gratia probe data is making it into the database at
rsv.grid.iu.edu?

One small hitch though -- the init script for osg-rsv includes

##
## Source the Condor-Devel setup file
##
. $VDT_LOCATION/vdt/etc/condor-devel-env.sh


I ran the configure_osg_rsv script after setting
VDTSETUP_CONDOR_DEVEL_LOCATION = /home/condor (base of the 7.0
installation), but this line still refers to the condor_devel in
$VDT_LOCATION. I had to comment that out of course since we are not
using the VDT condor-devel. The hitch is that vdt-control regenerates
the init scripts each time it enables a service, and thus this change is
lost at each "vdt --off osg-rsv; vdt --on osg-rsv" cycle. Do you know
how I can permanently get rid of this line?

Thank you for the assistance,

-Wayne



Scot Kronenfeld via RT wrote:
> I ran a VDT 1.8.1 test locally using a pre-existing Condor 7 installation
> for both Condor and Condor-Devel. The RSV jobs and some manual jobs I
> submitted all ran fine.
>
> During the install, you usually set VDTSETUP_CONDOR_LOCATION to point at
> your pre-existing install. When installing OSG-RSV, you will also need to
> set VDTSETUP_CONDOR_DEVEL_LOCATION to point at this same install.
>
> Let me know if you have any questions.
>
>
>
CC: Arvind Gopu <agopu@indiana.edu>
Subject: Re: [vdt-support #3258] RSV problem at STAR-BNL OSG site
Date: Wed, 30 Jan 2008 16:12:26 -0600
To: vdt-support@OPENSCIENCEGRID.ORG
From: Scot Kronenfeld <kronenfe@cs.wisc.edu>
Download (untitled) / with headers
text/plain 755b
Hi Wayne,

> After upgrading condor to 7.0 and configuring it to allow 'local'
> universe jobs, the RSV jobs appear to be running successfully. Can you
> check to see if the Gratia probe data is making it into the database at
> rsv.grid.iu.edu?

Arvind will have to check this. I've CCed him on the email.

> One small hitch though -- the init script for osg-rsv includes
>
> ##
> ## Source the Condor-Devel setup file
> ##
> . $VDT_LOCATION/vdt/etc/condor-devel-env.sh

The condor-devel-env.sh file should exist, and be pointing at
/home/condor. It should set CONDOR_CONFIG and CONDOR_DEVEL_LOCATION
to their correct values. It may not be necessary, but it shouldn't
hurt either :)

Let me know if the values in that file look correct.

Thanks,
Scot
CC: Scot Kronenfeld <kronenfe@cs.wisc.edu>
Subject: Re: [vdt-support #3258] RSV problem at STAR-BNL OSG site
Date: Wed, 30 Jan 2008 18:08:08 -0500 (EST)
To: vdt-support@OPENSCIENCEGRID.ORG
From: Arvind Gopu <agopu@indiana.edu>
Download (untitled) / with headers
text/plain 1.4k
Hi Wayne,

Yes, your STAR-BNL resource appears to be uploading RSV records
consistently for the past several hours.

As a sidenote, if you've not done so, feel free to configure your VDT
apache to serve RSV pages as explained here.
http://rsv.grid.iu.edu/documentation/vdt-package.html#configure_osg_rsv_script_localweb

If you did that, you'll also be able to view your RSV results locally at
https://stargrid02.rcf.bnl.gov:8443/rsv

And btw, ignore the directory-CE-permissions probe warnings - we're
working on a better version of that probe. I do notice expired CRLs though
that you might want to fix :-)

Cheers,

Arvind


On 2008-01-30 16:12 (-0600), Scot Kronenfeld had pondered:

> Hi Wayne,
>
> > After upgrading condor to 7.0 and configuring it to allow 'local'
> > universe jobs, the RSV jobs appear to be running successfully. Can you
> > check to see if the Gratia probe data is making it into the database at
> > rsv.grid.iu.edu?
>
> Arvind will have to check this. I've CCed him on the email.
>
> > One small hitch though -- the init script for osg-rsv includes
> >
> > ##
> > ## Source the Condor-Devel setup file
> > ##
> > . $VDT_LOCATION/vdt/etc/condor-devel-env.sh
>
> The condor-devel-env.sh file should exist, and be pointing at
> /home/condor. It should set CONDOR_CONFIG and CONDOR_DEVEL_LOCATION
> to their correct values. It may not be necessary, but it shouldn't
> hurt either :)
>
> Let me know if the values in that file look correct.
>
> Thanks,
> Scot
>
Download (untitled) / with headers
text/plain 502b
From John:
In response to my question to see if the su -c command is setting the environment
incorrectly.

[root@mstr1 ~]# . $VDT_LOCATION/setup.sh
[root@mstr1 ~]# . /home/osg/vdt/etc/condor-devel-env.sh
[root@mstr1 ~]# su -c "env" osg-rsv | grep CONDOR_CONFIG
VDTSETUP_CONDOR_CONFIG=/home/condor/condor_config
CONDOR_CONFIG=/home/condor/condor_config

which is my production CONDOR

i do set CONDOR_CONFIG in /etc/csh.cshrc which is osg-rsv's default
shell...could that have anything to do with it?
Download (untitled) / with headers
text/plain 526b
> [root@mstr1 ~]# . $VDT_LOCATION/setup.sh
> [root@mstr1 ~]# . /home/osg/vdt/etc/condor-devel-env.sh
> [root@mstr1 ~]# su -c "env" osg-rsv | grep CONDOR_CONFIG
> VDTSETUP_CONDOR_CONFIG=/home/condor/condor_config
> CONDOR_CONFIG=/home/condor/condor_config
>
> which is my production CONDOR
>
> i do set CONDOR_CONFIG in /etc/csh.cshrc which is osg-rsv's default
> shell...could that have anything to do with it?

This explains the problems that Wayne was having. We need to edit the RSV
documentation to mention this issue.