Biochemistry laboratories - 201
Jean-Yves Sgro -jsgro@wisc.edu
*Note*: To see as slides click on **"Gift icon"** at the top right. To revert back to page view click on **"book icon"** when on the first slide.
Terminal
return
keyThe Biochemistry Computational Cluster (BCC) is a High Throughput Computing (HTC) environment within the UW-Madison Biochemistry Department.
submit_node
|
|-[2TB disk]
-------------------------
| | | | | | | | |
N1 N2 N3 N4 N5 N6 N7 N8 N9
HTCondor is a “scheduler” system that dispatches compute jobs to 1 or more "compute nodes."
Using HTCondor is the only approved method for performing high throughput computing on the BCC Linux cluster.
Jobs have to be ready to be processed by HTCondor as jobs cannot be interactive on the cluster.
ssh myname@submit.biochem.wisc.edu
Submit a "job file" that contains information on how to run the software
myjob.sub
myjob.sh
Information about software and file transfer
Invert the process:
Login - [replace myname with your login name.]
ssh myname@submit.biochem.wisc.edu
Then move to /scratch
and create a directory with your name and another directory within to work with.
$ cd /scratch
$ mkdir myname #replace myname with e.g. YOUR ID
$ mkdir quickstart
$ cd myname/quickstart
Create execute file (e.g. with nano
)
$ nano hello.sh
File content:
#!/bin/sh
echo "Hello World"
Make file executable:
$ chmod u+x hello.sh
$ nano hello.sub
File content:
executable = hello.sh
should_transfer_files = Yes
output = hello.out
error = hello.err
log = hello.log
queue
Resource | Link |
---|---|
HTCondor Quick Start Guide | http://research.cs.wisc.edu/htcondor/manual/quickstart.html |
Complete manual* | http://research.cs.wisc.edu/htcondor/manual/ |
*You can check which manual you need by checking which version of HTCondor is installed with command: condor_version