Powered by:
Open Science Grid
Center for High Throughput Computing

Compiling or Testing Code with an Interactive Job

The CHTC has two "build" machines configured for compiling code interactively. Furthermore, any CHTC execute server can be used for testing commands and programs before submitting large batches of non-interactive jobs ('typical' jobs like our "hello-world" example).

Typically, any compilable code to be run in CHTC should be compiled on our compiling servers.

(Note: If you are using a non-CHTC submit server to submit jobs to a non-CHTC HTCondor Pool, you may have different capabilities for submitting jobs interactively that are not covered here.)

1. Setup

Though it's not essential, it may be easiest for you to create a tarball with all necessary code and compiling sources/libraries that you'll need for the interactive job, so that you don't have to list every file and/or directory separately for "transfer_input_files" in the submit file.

For example, you could use something like the following command within a directory where all of your needed files and directories are located:

[alice@submit]$ tar -czvf combined.tar.gz file1 file2 dir1 (etc.)
The above would produce a single file called "combined.tar.gz" that contains all other files and directories indicated at the end of the command.

2. Creating Interactive Submit Files

Create a file called interactive.sub (or any other name you want), and paste in the below lines:

universe = vanilla
# Name the log file:
log = interactive.log

# Name the files where standard output and error should be saved:
output = process.out
error = process.err

# If you wish to compile code, you'll need the below lines. 
#  Otherwise, LEAVE THEM OUT if you just want to interactively test!
+IsBuildJob = true
requirements = (OpSysAndVer =?= "SL6") && ( IsBuildSlot == true )

# Indicate all files that need to go into the interactive job session,
#  including any tar files that you prepared:
transfer_input_files = combined.tar.gz, other_file.txt

# It's still important to request enough computing resources. The below 
#  values are a good starting point, but consider your file sizes for an
#  estimate of "disk" and use any other information you might have
#  for "memory" and/or "cpus".
request_cpus = 1
request_memory = 1GB
request_disk = 1GB

queue
Modify the filenames for output, error, and log per your preferences.
Modify the filename for transfer_input_files to indicate the source tarball you created.

For SL6 submission
The above submit file will select for our Linux 6 (SL6- and RHEL6-compatible) build machines by default.

3. Submitting and Working Interactively

When you are ready to go, enter the following with your correct submit file indicated instead:

[alice@submit]$ condor_submit -i interactive.sub
After several minutes (the time it takes for HTCondor to match new jobs to available CHTC execute servers), an SSH connection will be created to an execute server, and the relevaant files will have been transfered into the job working directory for you. From there, you can un-tar any tar files and execute and compiling or test commands as you please.

Typical un-tar command:
[alice@build]$ tar -xzvf combined.tar

When you are done working, make a new tarball of files you want copied back to the submit server, and delete everything else that you don't want.

4. Exiting

To exit as you would for any SSH shell, type exit. Your interactive job will be terminated, and you will return to the location on the submit node from which you submitted the interactive job. Any new files (your new tarball) will have been copied back to this location.