Sysadmin Task List

• Storage server deployment.
  • Wipe old BeeGFS server.
  • Set up rsync server.
    • Make sure /cache is also backed up via rsync
  • Set up new tape backup.
  • Set up new cloud backup (Azure via IT).
  • Erase old cloud backup.
• Compute node deployment to C4. Depends on supply chain / buyer.
  • Update ship-to address when PO gets executed.
  • Secure delivery of node
  • Decommission n20 -- do we want to do something with it?
• New VMWare server. Replaces current servers (too old to be updated). Could be faster CPUs, more RAM, and get hulao redundancy.
  • Requires new network switches to get Swarm private network. Waiting on Super Micro.
• Add gnomon.ucsf.edu to our SSL cert
• Enable scrontab or figure out where to schedule cron jobs on the cluster

Things for us to know:

• Have to renew VMWare licenses every year -- like $200, but hard to tie to a PO. Used CC and got reimbursed.
• We need to upgrade CentOS systems to, probably, RockyLinux 8 (to stay in lock-step with Wynton)
• Could submit Slurm jobs from worker nodes (would need configuration)

Obscure questions:

1. How is the tape database itself backed up/restored?
2. Is there an API to the tape database to query file status?