Linux Working Group

AGENDA LOCKED

Meeting Date

• TEAMS: 2022-05-04

Invitees - Attendees

• Invited: Adrian, Anthony (group leader), Clayton, Guoxiang, Lori, Fraser, Devon, Nathan, Nick, Todd, Dave, Lawrence, Omar
• Present: Adrian, Anthony (group leader), Clayton, Guoxiang, Lori, Fraser, Devon, Nathan, Todd, Lawrence, Omar

Review and accept previous meeting minutes.

• CsLWGMeeting20220420

Review last meeting's Action Items

• Anthony - will create a ticket for monitoring processor data
  • Devon and Anthony making progress
  • Combined general use cpu load graph (and possibly other metrics)
• Clayton/Fraser - document process of adding hosts to AD and move to a generally accessible place (git) -> https://rt.uwaterloo.ca/Ticket/Display.html?id=1217894
  • waiting for Fraser to test it out on graphics lab machines
• Dave - put up Beta version of new Virtual Host Index / Anthony to create a ticket - RT#1211603 -> working on it
  • not up yet
• Lori - create ticket for Devon to create combined graph for DFSc https://rt.uwaterloo.ca/Ticket/Display.html?id=1215768
  • ticket created for network bandwidth (RT#1215768)
    • fixed issue with averaging
  • still need to create ticket for "combined graph" - load? CPU? across all machines?
    • may need some definitions, but generally shows overall health of the Ceph cluster
  • Devon will create a combined graph of some type - can be refined
    • Total DFSc CPU monitoring usage added at the top of https://icinga.cscf.uwaterloo.ca/grafana/d/03EnhXZGz/dfsc-monitoring?orgId=1
• Lawrence - follow-up with SuperMicro re: RT#1079451 - ubuntu1804-006 CPU hardware errors
  • Looks like CPU is in flight
    • need to verify that - not clear to me - LF will follow-up
• Anthony - update the Purpose field of currently "unused" machines in the Virtual Host Index: some progress, need to make another pass.
  • progress was made, some more work to do

Proposed Agenda Items

• proposed reboots coming up soon?
  • ideally Thursday after 5pm, will take a couple hours
  • machines/services involved? all (most) cloud machines - general use servers - (ubuntu1804*, ubuntu2004*)
    • vault and web services should be unaffected as machines will be rebooted one-at-a-time
    • *411 hosts outside of scope of this reboot, so MySQL services will be unaffected (and dependent apps)
  • Lawrence/Omar will send out a notice
  • again suggest updating all motd files - Anthony will do ASAP
  • kernel will be reverted to 5.4 on most machines (can be put back to 5.13, if necessary)
• Consider btrfs usage? Testing/benchmarking. Positives: snapshots, compression. RAID0/1, volume mgmt. Negatives: write hole/RAID5/6, code churn, stability - Nathan
  • do we use parity RAID? May not be an issue
  • tabled for next meeting

Action Items for next meeting

• Anthony - create dashboards for monitoring processor data
• Clayton/Fraser - document process of adding hosts to AD and move to a generally accessible place (git) -> https://rt.uwaterloo.ca/Ticket/Display.html?id=1217894 - Fraser to test process on GL machines
• Dave - put up Beta version of new Virtual Host Index / Anthony to create a ticket - RT#1211603 -> working on it
• Lori - create ticket for Devon to create combined graph for DFSc
• Lawrence - follow-up with SuperMicro re: RT#1079451 - ubuntu1804-006 CPU hardware errors - verify CPU being shipped
• Anthony - update the Purpose field of currently "unused" machines in the Virtual Host Index: some progress, need to make another pass.
• Nathan - create ticket for btrf testing, especially wrt for managing LXC root file size limits
• Lawrence/Omar - send out notice regarding upcoming between-term reboots, Anthony - update relevant motd files