Hi All,

Thanks to this email[1] from Peter Cock I was able to find what was
causing the error I initially described here[2]. I also now have a
theory of why this is happening.

My unconfirmed guess. When a colon is present in the standard output
or standard error file, someone betweent Galaxy, python-drmaa,
pbs-drmaa, Torque or Maui recognize this as a scp syntax, like:
user@host:/path/to/file

And then choose to at 'galaxy@' in from of what is really a filesystem
path. This is why I'm seeing this kind of error:
Unable to copy file /var/spool/torque/spool/1013.head.OU to
galaxy@/local/opt/galaxy/galaxy-dist.8082/database/job_working_directory/000/9
/task_0/9:1.drmout
*** error from copy
cp: cannot create regular file
`galaxy@/local/opt/galaxy/galaxy-dist.8082/database/job_working_directory/000/9/task_0/9:1.drmout':
No such file or directory
*** end error output
Output retained on that host in: /var/spool/torque/undelivered/1013.head.OU

By using Peter's ".replace(":", "_")" trick on
"lib/galaxy/jobs/runners/drmaa.py" I was able to make this issue go
away. As Peter mention this is a very dirty fix and I'm sure there has
to be a much cleaner way of fixing this issue.

[1]http://dev.list.galaxyproject.org/Galaxy-with-Univa-Grid-Engine-UGE-instead-of-SGE-td4657848.html
[2]http://dev.list.galaxyproject.org/Issue-when-enabling-use-tasked-jobs-with-torque-and-nfs-td4657294.html

I'll be opening a ticket on Trello to follow the progress on this issue.
Regards,
Carlos
___________________________________________________________
Please keep all replies on the list by using "reply all"
in your mail client.  To manage your subscriptions to this
and other Galaxy lists, please use the interface at:

  http://lists.bx.psu.edu/

Reply via email to