jamierajewski/illume-v2

Issues

• Add interactive cpu worker capability

  #62 opened 3 years ago by jamierajewski
  0

• Make shared-data.tf DRY

  #44 opened 3 years ago by jamierajewski
  1

• Fix worker config to mount the ephemeral storage correctly

  #60 opened 3 years ago by jamierajewski
  3

• Mount /etc/ssh from NFS

  #59 opened 4 years ago by jamierajewski
  0

• Running the STRAW paper code in Jupyter doesn't work - missing LaTeX component(s)?

  #56 opened 4 years ago by jamierajewski
  1

• Figure out domain name

  #33 opened 4 years ago by jamierajewski
  1

• Revisit issue of IP's changing on a deployment

  #51 opened 4 years ago by jamierajewski
  2

• Create volume for Condor to allow the queue to persist during updates

  #58 opened 4 years ago by jamierajewski
  1

• Track hashes of templates so that Terraform recognizes a change

  #55 opened 4 years ago by jamierajewski
  1

• NFS mounts don't always work on boot via /etc/fstab

  #57 opened 4 years ago by jamierajewski
  1

• Add alerts for certain issues in Prometheus/Grafana

  #29 opened 4 years ago by jamierajewski
  0

• CVMFS mounts are missing

  #50 opened 4 years ago by jamierajewski
  1

• Set up JupyterLab with HTCondor

  #54 opened 4 years ago by jamierajewski
  0

• Create ingress /scratch policy

  #53 opened 4 years ago by jamierajewski
  0

• Make login node's /scratch accessible by worker nodes

  #52 opened 4 years ago by jamierajewski
  0

• Condor Interactive jobs not working

  #24 opened 4 years ago by jamierajewski
  1

• Condor tries to run interactive jobs as 'nobody' causing failures

  #45 opened 4 years ago by jamierajewski
  1

• Deal with important volumes in a generic way (no NFS)

  #49 opened 4 years ago by jamierajewski
  0

• Rust seems to be installed locally, not globally

  #46 opened 4 years ago by jamierajewski
  1

• Find a smoother redeployment strategy

  #48 opened 4 years ago by jamierajewski
  1

• Set up automatic snapshots of important volumes on an interval

  #47 opened 4 years ago by jamierajewski
  0

• LDAP currently only sets up from a backup

  #43 opened 4 years ago by jamierajewski
  0

• How to handle updates?

  #20 opened 4 years ago by jamierajewski
  5

• Harden network

  #15 opened 4 years ago by jamierajewski
  2

• Reboot GPU workers after cloud-init finishes

  #42 opened 4 years ago by jamierajewski
  1

• Cut down CUDA stuff

  #40 opened 4 years ago by jamierajewski
  1

• Add more channels to Conda

  #28 opened 4 years ago by jamierajewski
  2

• Add user-accessible Grafana dashboard for job monitoring

  #34 opened 4 years ago by jamierajewski
  1

• Create build scripts

  #22 opened 4 years ago by jamierajewski
  2

• Use token auth for Condor

  #18 opened 4 years ago by jamierajewski
  0

• Set static IP's for applicable nodes

  #14 opened 4 years ago by jamierajewski
  1

• Explain what Illume is in the readme

  #36 opened 4 years ago by jamierajewski
  1

• Handle secrets better

  #41 opened 4 years ago by jamierajewski
  0

• License?

  #37 opened 4 years ago by jamierajewski
  2

• Ensure that ephemeral storage is properly used in deployments

  #30 opened 4 years ago by jamierajewski
  0

• Test fresh repo

  #39 opened 4 years ago by jamierajewski
  0

• Add secgroups to terraform

  #38 opened 4 years ago by jamierajewski
  0

• Organize the docs

  #35 opened 4 years ago by jamierajewski
  1

• Convert Packer JSON to HCL2 format

  #32 opened 4 years ago by jamierajewski
  1

• Relabel hosts in Grafana monitoring to match the OpenStack instance names

  #31 opened 4 years ago by jamierajewski
  1

• Create custom welcome message

  #21 opened 4 years ago by jamierajewski
  1

• Figure out why Podman + GPU pass-through doesn't appear to work anymore

  #25 opened 4 years ago by jamierajewski
  0

• Why GPU jobs won't distribute to all GPU nodes

  #26 opened 4 years ago by jamierajewski
  2

• Create important volumes ahead of time, and attach them at deploy

  #23 opened 4 years ago by jamierajewski
  1

• Nvidia-smi doesn't run through Condor

  #27 opened 4 years ago by jamierajewski
  1

• Add dedicated interactive-only worker nodes

  #13 opened 4 years ago by jamierajewski
  1

• How to handle passwords during deployment?

  #16 opened 4 years ago by jamierajewski
  1

• Use Ansible in place of remote-exec?

  #17 opened 4 years ago by jamierajewski
  1

• Refactor remote-exec inline code into scripts

  #19 opened 4 years ago by jamierajewski
  2

• Add cgroup limits on condor jobs

  #12 opened 4 years ago by jamierajewski
  0