EngineeringUnits/Management/Meetings
<< <  2020 / 8 >  >>
Mon Tue Wed Thu Fri Sat Sun
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31            

Techleads meetings: Mondays 5pm UTC and Thursdays 9:30am UTC

Agenda

Owner

Topic

your name here

add your own

Joey/James W

Panda porter boards in the DC

Michael Hope

PowerPC toolchain bugs

Michael

Building Ubuntu ourselves

scottb

Hardware pack development and Landing Teams

Actions from previous meeting

Attendees

  • Andy Green - present
  • Angus Ainslie - present
  • Amit Kucheria - present
  • David Rusling - traveling
  • James Westby - present
  • Jamie Bennett - present
  • Michael Hope - present
  • Alexander Sack - vacation
  • Christian Reis - traveling
  • Jesse Barker - present
  • Joey Stanford - present
  • Lee Jones - absent
  • Loïc Minier - present, tardy
  • Paul Larson - present
  • Paul McKenney - absent

  • Sachin Gupta - excused
  • Scott Bambrough - present
  • Stephen Doel - traveling
  • Steve Langasek - present

Minutes

  • Quick discussion about the panda boards. They are almost ready but not quite yet. Target is Friday.
  • PPC toolchain bugs
    • bug found in the toolchain building the kernel for ppc in Ubuntu
    • it's way out there in terms of us providing support
    • we can't reproduce. we don't have the hardware. we don't want the hardware
    • it's not related to arm
    • discussion surrounding how to deal with this
    • Loic believes that it's low priority to fix this
    • MH is thinking that having a plain vanilla Linaro deb pkg to compare against would help
  • Building Ubuntu ourselves
    • James W thinks that if David M can get his builders up and running then there might be enough capacity there to rebuild the entire archive.
    • Loic thinks we need native hardware to do rebuilds
    • MH found qmeu 3 times slower than the clock speed
    • Andy reported Fedora had the same findings
    • This effort could help us find regressions ourselves and thus, before someone else finds them
    • James W mentioned the OBS guys might be helpful.
  • HWPACK and Landing Teams
    • The amount of hoops we need to go through to build one and get it maintained is excessive. Can we streamline it all?
    • discussion surrounding how hwpacks are created
    • no automatic way to attach hwpacks at the moment
    • Loic said that no steps are really hard but some steps are a little trying, e.g. pinging IS for everything
    • Angus said that once it's on snapshots it's a lot less painful but getting it on snapshots via IS takes a week (24 to 36 hours via IS but a week for Linaro due to timezones)
    • James W said that once our own instance (not the one we share with OEM) is up these issues will for the most part disappear.
    • Offspring HW is awaiting setup in London and Joey expects it to up by the 18th.
    • James W said there is some offspring coding needed to be able to automate this better for Linaro.
    • James W thinks the main problem is in the communications and exceptions between the landing teams and the platform.
    • Loic said that we should be escalate after 2 days vs 9 days
    • Scott & James W to take this offline and find solution

    • Jamie mentioned that we have a lot of hwpacks and we should think about consolidating them. Scott replied that George thinks we should have only four.

Actions from this meeting

  • Scott & James W to talk about hwpacks and offsprint and report back with recommendation on improvements.

EngineeringUnits/Management/Meetings/2011-03-10 (last modified 2011-03-10 18:42:56)