From MozillaWiki
Jump to: navigation, search
  • Time: Tuesday at 9:15 AM PST / 12:15 PM EST / 5:15 PM UTC.
  • Place: Mozilla HQ, Warp Core
  • Phone (US/Intl): 650 903 0800 x92 Conf: 8616#
  • Phone (Toronto): 416 848 3114 x92 Conf: 8616#
  • Phone (US): 800 707 2533 (pin 369) Conf: 8616#



  • SCL status
    • Py out, PHP in
    • DB deadlock preventer
    • Capacity status
      • We can spool up enough machines to *at least* handle all of what's in PHX now.
      • 72 DBs total, 5 live right now.
      • Still capacity testing (next to prod)
  • Reminder -- Ops is in "lock down" until at least 24 Mar 2011


Python Sync Server (Tarek)

Python Reg Server (Tarek)

PHP Sync Server (Toby)

  • Reverting to PHP in SCL for FF4 launch
  • Working on infra scripts that were abandoned due to Python changeover
  • Indexes in production do not match the ones in stage or documented in code. This includes PHX (SCL ones were way off)
    • Unclear if this is a problem - is cheaper write load offsetting lack of a range index?

Identity Server (JR/rnewman)



  • Continue work on M1 tasks.
  • Continue work with rnewman re: JS Lib interface.

Firefox Home (Stefan)

Sync Client (Philipp)

  • Last Fennec blockers landed
  • Released Sync 1.7 add-on to match
  • Fleshed out Sync engineering priorities post Firefox 4

Notifications (Shane/Alex)

  • End-to-end crypto fully functional
  • Working on harnessing test framework for client-side code
  • Waiting on ops to set up a VM to use as a demo server
  • Building demo web apps to run on VM once set up


  • Tested PHP scl2 integration. Found issue around tabs not syncing due to memcache. Fixed and signed off
  • Working on a Sync testing helpful tips video for QMO.
  • Let us know if you need testing for anything in particular this week.






Notes and actions

  • Python was holding open a lot of mysql connections and there were index issues.
  • Action: Jeff to drive a post-mortem and figure out next steps during the work week as Tarek will also be here.
  • Let's do a complete end to end user test for when we migrate a user to make sure all of their data comes through (some reports of Tabs not coming back and needing a memcache flush).
  • Action: Jeff will file a bug to track this.
  • The database index (time based) do not match up between staging and production. This has impact on capacity, but should not cause the servers to meltdown. We are still highly confident that this should not impact launch.
  • Action: Pete is going to run some load tests and figure out next steps.
  • Identity M1: target date early April.
  • Action: JR to reach out to clyon's team to do a preliminary security assessment on the protocol and budget in time for code reviews.
  • MConnor and Stuart talked last week and he hasn't heard anything from Stuart.

Follow ups from last week

  • By next meeting we should have a project plan for Identity ready for review.
  • Fennec Perf - we have done everything we can from Sync's end. Other fixes will need changes to APIs Sync uses (like Password Manager, Places etc).
    • MConnor to follow up with Stuart to ensure we're all on the same page.
  • 1.7 add-on QA sign off targeting tomorrow (assuming no major issues). Tested and Shipped
  • Jeff and Tony to figure out sign off process. Worked it out with team, more notification upfront

Fx 4 client

Fx 4 server

Other issues