*Attendees*
Eric, Jean, Ginnie, Barbara, Jan, Tim, Clay, Alok, Evan, Ethan, Dana,
Karen, Sundar, Sue, Mary, Joe, William, Moriah, Tycho, Jack
*Notes*
>
> Thanks to Sundar for running last week's meeting while I was out.
First, thanks to everyone on your progress. We're looking good.
>
> Agenda:
>
> Projects status:
> - AI
- Sundar and Tycho tried to boot AI image; can now use wanboot and http.
- At first, had to use vendor-specific options with sparc which
Sundar wants to avoid.
- Can download but image panics when try to boot.
- Seems to be a bootroot problem related to network and to the
T1000, as:
- Karen booted from harddrive OK, Sundar using network.
- Jan can boot on T5220, but problem specific to T1000.
- Sundar verified that a stock Nevada bootroot works on T1000.
- Sundar to try on sun4u.
- Once this issue is figured out, Sue can automate the fix in installadm.
- William: orchestrator and target discovery changes:
- Most changes are skipping over fdisk changes.
- Will take three days to code/unit test.
- William go on vacation soon, will give to Sundar / Alok to
finish if needed.
> - SPARC team (Greg, Tycho)
Thanks for initial SVR4 package work. It was needed as a backup
- Tycho worked on getting wanboot to work with Sundar.
- Now switching gears to ICT changes. Making progress.
- libti and disk labeling will come later; not impacting AI team as yet.
> - DC
- Jean is cleaning up the code. Sent preliminary review to Tycho and Jan.
- Added 20Mb extra space to bootroot to allow tweeking of the bootroot
without building a new image.
> - Snap
- Done with sparc changes.
- Jason's done with testing. Ported test suite to SPARC.
>
> Machine update:
> - sun4v development/test machines and ldoms
- The goal is to have a build "machine" (ldom virtual machine is OK) for
each internal developer for development/testing.
- Clay has worked at setting up line2-t1000 for ldoms. Found docs for
S10, none for NV.
- Tycho to find pointer to ldoms docs for NV.
- Docs and NV ldoms setup recipe to be made available internally.
> - build machine (osol-bldsp.sfbay)
- A V480 will be renamed to osol-bldsp.sfbay (lab request has been
submitted).
- System has 4 CPUs and 16Gb RAM.
- Mary is looking further into getting larger disks than the 2 72Gb
disks it already has.
>
> Testing:
> - Plans and status by project (DC, Snap, AI)
Snap:
- already reported.
DC:
- That AI can boot the image generated by DC is proof that DC is working.
- Will enlist Erik Lafever (did 2008.11 DC testing) to test (incl
regression test).
AI:
- Andre M. and Jeffrey H. will test SPARC when available.
- Will be automated. Any testing done now is manual.
General:
- QE needs to be engaged to automate testing for DC and everything else.
- Joe has been looking into long range regression testing.
- As more functionality / platforms come in to the fold, the test
matrix increases.
- The challenge is to figure out how to set things up so that when
an install bug gets fixed, how to regression test on whole matrix. This
is to reduce the burden of testing the whole matrix when a developer
fixes a bug.
>
> Roundtable
Nothing was brought up.
Thanks,
Jack
