[Swift-user] Running first.swift remotely on NCSA

[Andriy Fedorov]

>
> I will have to figure it out...
>

The problem turned out to be in the setup of our firewall. For the
time being, I cannot resolve it, so I am trying to use TeraGrid UC
site for testing.

Unfortunately, I ran into problems again.

Here is my sites.xml:

<config>

<pool handle="UC-GT2-Fork">
        <gridftp url="gsiftp://tg-gridftp.uc.teragrid.org" />
        <jobmanager universe="vanilla"
url="tg-grid.uc.teragrid.org/jobmanager-fork" major="2" />
        <workdirectory>/home/fedorov/scratch</workdirectory>
</pool>

<pool handle="UC-GT2-PBS">
        <gridftp url="gsiftp://tg-gridftp.uc.teragrid.org" />
        <jobmanager universe="vanilla"
url="tg-grid.uc.teragrid.org/jobmanager-pbs" major="2" />
        <workdirectory>/home/fedorov/scratch</workdirectory>
</pool>

<pool handle="UC-GT4">
        <gridftp url="gsiftp://tg-gridftp.uc.teragrid.org" />
        <execution provider="gt4" jobmanager="PBS"
        url="https://tg-grid.uc.teragrid.org:8443/wsrf/services/ManagedJobFactoryService"
/>
        <workdirectory>/home/fedorov/scratch</workdirectory>
</pool>

</config>

And here is tc.data:

UC-GT2-Fork     echo_gt2_fork   /bin/echo       INSTALLED
INTEL32::LINUX  null
UC-GT2-PBS      echo_gt2_pbs    /bin/echo       INSTALLED
INTEL32::LINUX  null
UC-GT4          echo_gt4        /bin/echo       INSTALLED
INTEL32::LINUX  null

I am trying to run example first.swift, but use echo_gt2_fork,
echo_gt2_pbs, and echo_gt4 instead, to test different setups.

I have success with echo_gt2_fork and echo_gt4. But when I try to use
echo_gt2_pbs, it never finishes, and I see this:

[fedorov at tg-login1 vdsk]$ swift first.swift
Unable to find required classes (javax.activation.DataHandler and
javax.mail.internet.MimeMultipart). Attachment support is disabled.
Swift 0.5 swift-r1783 cog-r1962

RunID: 20080620-1514-in8cf2yg
Progress:
echo_gt2_pbs started
Progress:  Stage in:1
Progress:  Executing:1
Progress:  Executing:1
...

Note, that I can see my job started and completed with "qstat" on UC
site, but the result never gets back. This is on
tg-login.uc.teragrid.org, so there should be no problem with firewall.

The only reason I would like to get this working for pre-WS is because
my true goal is to be able to run MPI job, and be able to pass node
type to PBS. The only way to specify host type for GT4 GRAM is through
Job description extensions (see
http://www.globus.org/toolkit/docs/4.0/execution/wsgram/WS_GRAM_Job_Desc_Extensions.html),
and I don't know how this can be translated into tc.data. With GT2, I
can simply use "host_count=2:compute".

Can someone help me with this simple example, so that I can try to
make more complicated scripts work and run MPI applications?

Thank you

Andrey