Home

Lauren, Aaron, Adam

1. Remote Connections and File Transfers
   • tasks: ssh, scp, tar
2. Introduction to the Cluster, Queue, and Partitions/Allocations
   • tasks: view the queue, view list of servers, view partition/allocation info
3. Submit and Monitor a Job
   • tasks: write a submit file, submit it, view it the queue, make sense of log/out/err files
4. Interactive Sessions
   • tasks: run an command-line job through the scheduler, run a GUI tool within an interactive session
5. Software and MPI
   • tasks: view modules or software installed; submit a job using installed software; install your own software, submit a job that depends on MPI
6. File Handling
   • tasks: submit jobs with each of 2-3 file availability methods (different file systems, etc.) (still in progress)

Ron, Anita, Plamen

1. batch scripts/queues
2. monitoring jobs use of resources - cpu, memory, I/O
3. parallel/ concurrent serial strategies/workflows
4. data movement (rsync, globus, tar, etc)
5. evaluating program scaling
6. diagnosing job failures

Ista, Martin, Patricia:

1. Accessing the cluster/transferring data to the cluster
2. Understanding the queue/viewing the status of your job
3. Submitting jobs
4. Understanding types of parallelism
5. Writing a shell script that launches independent parallel jobs

Sean, Radhika, Andrew

0. Introduction to how cluster is structured, login nodes, backend nodes, job scheduler. (explain the jargon before diving in head first)
1. SSH, file system (log in, ls/df filesystem building on unix shell lessons)
2. file transfer, scp, rsync, globus (transfer files and download files to laptop, setup globus account & transfer files)
3. interactive session / debugging, software modules (run interactive job, load modules, queue/partition view)
4. batch job submission, job monitoring (batch script creation, submit batch job)
5. job array(create job array script and submit, review output)
6. visualization IPython, RStudio, Visit...(run interactive session and launch visual tools)

Bob, Amir, Chris

1. Logging in/SSH keys
   • login via ssh, keygen, copying keys, scp
2. Submitting jobs & understanding queues/resources
   • submit batch job
   • submit interactive job
   • submit jobs to different queues
   • submit jobs using doifferent resources
3. Job control
   • submit exmplaes
   • kill jobs
   • get job status
   • modify jobs
   • hold/resume
4. MPI & multithreaded
   • submit jobs having different shapes
   • talk about run diufference on clustser
5. for loops / job arrys
   • submit jobs using bash for loop
   • submit jobs via job arrays
6. job dependencies
   • submit jobs using dependency conditions
   • ADV: submit job array dependent arrays

Sean, Troy, Simo:

1. AWS - half hour intro, description of nodes and clusters, and set-up of AWS account, possibly with preinstalled software so all the students, regardless of their home institution, have access to the same kind of environment / topology,login with SSH keys, key management....
2. Deploy code on AWS from where it was stored on git repository
3. Parallel workfolow (data preprocessing / preparation, planning of assignment of data subsets to parallel routines)
4. parallel processing in R (and possibly other packages) - extension of modular programming because MP is a way of organization that makes it easier
5. how to run the code - shell script that distributes the execution
6. How to have the parallel executing programs write their results to a single database

Christina, Avalon, Mary

1. "Intro" - what is a node, what is a cluster (diagrams!)
2. Logging into remote servers (probably ssh) -> activity: log into the training cluster
3. Dealing with files - where to store, access times, restrictions of file size, backups -> activity: compare accessing a file on one system versus another
4. Queuing system -> activity: write a submission script (and submit?)
5. MPI concepts - what is a task, comparison to shared memory -> activity: write a basic MPI script?