[Swift-devel] Problems running coaster

[Michael Wilde]

On 7/27/08 2:44 PM, Ben Clifford wrote:
> I don't see the original message for this so I can't see the logs.

It seemed to go through - I got the message from the list.

But if it didnt get to you, the files are on the CI net at 
~wilde/coast/crob1, and the text is below.

> The 
> list software used to filter messages with large attachments (and 
> hopefully still does).

the file was 10K bytes

> What cog version do you have?

2093

Something like r2066 fixes a bug with
> walltimes that was breaking coasters.
> 

-------- Original Message --------
Subject: [Swift-devel] Problems running coaster
Date: Sun, 27 Jul 2008 14:10:20 -0500
From: Michael Wilde <wilde at mcs.anl.gov>
To: swift-devel <swift-devel at ci.uchicago.edu>

I got errors trying coaster both on the abe site on teragrid and locally.

Im using swift rev 2148

For both, I see in the log a message like:

DEBUG vdl:execute2 APPLICATION_EXCEPTION jobid=echo-aa3vo3xi -
Application exception: Could not submit job

Do you see whats wrong here?  I will continue to debug in the meantime.

Info below.

Thanks,

- Mike



The abe log is *l4.log (letter L), the local one is *ha.log.

The local one got a bit further, in that I see in the log the GETs of
the jar files in the bootstrap process.

The Swift script is:

type file;

(file t) echo (string s) {
      app {
          echo "the string is" s stdout=@filename(t);
      }
}
file outfile <"echo_000.txt">;
string words[] = ["s000","s001","s002"];
outfile = echo(words[0]);

(testing one echo call before I try a loop)

The local sites entry is:

<pool handle="localhost">
      <gridftp  url="local://localhost" />
      <execution provider="coaster" jobmanager="local:local"
url="localhost" />
      <workdirectory >/home/wilde/swiftwork</workdirectory>
</pool>

The abe sites entry is:

<pool handle="abe" >
    <execution provider="coaster" url="grid-abe.ncsa.teragrid.org"
jobManager="gt2:pbs" />
    <profile namespace="karajan" key="jobThrottle">4</profile>
    <gridftp url="gsiftp://gridftp-abe.ncsa.teragrid.org"/>
    <workdirectory>/u/ac/wilde/swiftwork</workdirectory>
    <profile namespace="globus" key="project">TG-CCR080002N</profile>


<!--altworkdirectory>/cfs/scratch/users/wilde/swiftwork</altworkdirectory-->
    <!--ROUXprofile namespace="globus"
key="project">TG-MCA01S018</ROUXprofile-->
    <!--SwiftDACprofile namespace="globus"
key="project">TG-CCR080002N</SwiftDACprofile-->

</pool>

tc.data has:

localhost       echo            /bin/echo       INSTALLED
INTEL32::LINUX  null
...
abe     echo            /bin/echo       INSTALLED       INTEL32::LINUX  null

All the files and logs are attached.