The following setup is based on Jaime Frey and PIC[1]'s IT team prototype [2] which aimed to run HTCondor jobs inside worker nodes at Barcelona Supercomputing Center [3]
- This repository contains the
request_glideins.shscript which needs to run from an Edge node on Theta with external connectivity. - It will use the collector as CCB.
- There are two halves to this, a first part which takes care of the COBALT side of things and gets all directories ready and in place. The second part will run a Singularity container tailored specifically for the setup. Said container will run the HTCondor part of the Split/starter on the Edge node (given that installing HTCondor is not an option).
- Have an active user account on the Theta supercomputer
- The account must be associated with a project which should also have its own directory on the Lustre shared storage
- Have administrative access to an HTCondor pool which should include PASSWORD as an accepted authentication method
- Have access to the password file mentioned on the previous point
- Have login access to a machine within the supercomputer network with a simple Singularity installation
Clone this repository in your home area i.e /home/myuser and run the request_glideins.sh with the parameters needed, making sure to specify the -u flag followed by your username.
cd fnalhpc_startd
./request_glideins.sh -n 1 -q debug-flat-quad -u myuser
- Is this all going to run under my account? -> Totally fine by me, condor binaries do live in shared storage
[1] https://www.pic.es/areas/#lhc
[2] https://htcondor-wiki.cs.wisc.edu/index.cgi/wiki?p=RunCmsJobsAtBsc
[4] https://htcondor-wiki.cs.wisc.edu/index.cgi/wiki?p=BuildingHtcondorOnLinux