TPTP-PMC-20080213

Logistics

Attending: Oliver, Chris, AlexA, AlexN, Paul, Joanna

Any objections to last week's minutes?

• no

Does anyone have to leave early?

• no

4.4.1 discussion

Oliver asks if there are any stoppers to 4.4.1

Paul comments on test pass results

• ~60% completion
• Most missing results are platform related
  • Joanna comments that some of the tests need to be removed. (~dozen)
  • Some of them are actual bugs with significant fix required
  • A few of the tests are in process now including a few bugfix results

Test has had some issues with a BIRT defect. Problem should be resolved when new Europa platform build is pulled in.

Joanna is working on some Europa logistics

Oliver asks if we are on track for on time release?

• Joanna notes that all all planned defects are in
• Paul notes that test pass should be done. Now have a few weeks for regression activities
  • this week for stop ship defects
  • next week regression testing and sync to latest Europa base
  • Regression with Europa release candidate (until the 29th)

We need to decide whether we will do smoke test our a lighter touch test

• Joanna suggests a need for a full pass on all the Agent controller (and security) touching bits (where majority of defects were found)
  • shouldn't be too bad
• Monday we grab from latest Europa milestone. Leave it to project leads to decide what should be tested.
  • Paul will generate a smoke test pass report next week.
  • Leads can go into CVS and disable tests that they don't want to run so that they don't contaminate the report
  • Test project will be rerunning items related to secure/unsecure
  • Oliver asks Trace and Monitoring projects about their status.
• AlexN for Monitoring: Yes, only a bit of additional testing needed
  • remote logging... which involves AC

Oliver asks Profiling if all key issues are resolved.

• AlexA: Has not heard any new complaints.
  • Eugene was going to test build. Alex has not heard of new issues since then and assumes that all is fine.

4.5 discussion

How is 4.5 I5 going?

Test

• a third to a half thru testing.
• Continuing with testing today.
• Assuming will finish tommorrow.
• Only issues thusfar have been 2 blocking BIRT defects.
• Paul will know for sure end of day tommorrow but currently things look okay for iteration.

Platform

• Team has barely scratched surface.
• Another blocker showed up yesterday.
• 1am build today had another blocker.
• No way will complete pass before early next week.

Monitoring

• Should be able to complete by end of Friday.
• Started test pass and running tests yesterday.
• Found some defects (mostly patches that existed but weren't checked in causing failing in logging).
  • They have been checked in and will appear in today's driver.
• Think can finish test pass by end of day friday.
  • May have a few test buckets that don't make it.
  • Will know for sure around thursday

Trace Will finalize within a week.

• Will contact Matt for an update because Richard is on vacation.

Oliver notes that we will be quite close to the deadline

• How bad is it to miss a delivery?
• Paul notes that the actual date for Ganymede is the 18th so there is the weekend of flex.
  • Paul does not see a disclaimer on the site about consequences
  • Suggests that we drive full steam ahead dropping I5
  • Backup plan would be to redrop I4.

4.5 OpenSSL

Joanna noted that legal denied request to ship OpenSSL.

• A similar request made on earlier eclipse project was also denied.
• Joanna has asked for a followup with legal.
  • They apparantly have some pedigree concerns
  • Earlier, legal stated there would be no difference in mechanism for either external dependence or shipping dependence
    • This would leave us with JSSE only as a solution and no real path to alternate solution
    • Does not leave us legally exposed because use of OpenSSL was modular.

EPP Packaging

We discussed the WTP packaging.

• all realize that there is a bit of extra effort.
• If can arrange it where the time we steal is from the right thing (i.e., choose what bugs to not fix to ensure coverage of bundle)
• Does anyone disagree about WTP packaging
  • Joanna has concerns regarding extra workload required (lots of defects for 4.5). team already stretched. Taking on more might not be managable.
    • Joanna trying to get handle on exactly what the extra workload will be. Wants particularly to know about additional testing requirements (can barely test what we have now)
    • Joanna had also heard there may be additional documentation, testing, etc. wants to know exactly what that effort is.
    • Chris agrees that better understanding the costs are needed.
  • Oliver notes that there should only be a little to set it up, recurring burden on test
  • Asks if we can we drop some of our existing testing to cover WTP testing requirements.
    • Oliver to Joanna: are there items in our test buckets that we could sacrifice?
    • There was some discussion about whether Rational does additional testing and could assist here.
    • Oliver noted that Harm seemed to think that it wasn't really any true new work and more about timing existing work than adding new work.

If we bundle with WTP we can do all our testing/driving on that bundle.

• Tell lead users that they should to use that packaging

The short form of our discussion is that there needs to be someone to go talk to EPP/WTP to drill in and figure out what additional to be done.

• Oliver asks who would be best person to find out from Dave Williams exactly what is required.
  • Paul and Oliver can get together with Dave Williams to get more significant details.

Oliver thinks WTP is a good sales pitch

• downloaders of WTP will be using the profiler
• Will we be scared off by the additional defects that will probably result?
  • How do we balance support of WTP versus our existing bugfix pipeline
  • Will need to ensure that bugs get fixed and that we help community
  • Can we borrow from our bugfix pipeline to ensure we cover the community?
    • Chris notes that we can only borrow time to some extent
    • i.e., borrowing from hc that we are already borrowing from for something else
    • POG effort is already borrowing from HC to work on the bug pipeline
    • With POG we contain the borrowing because we have limited/phased lead users we pay attention to. How will this work when WTP bundle users report bugs.

Oliver talks a bit about contribution of large user base.

• e.g., bugs reported by a bunch of people should be higher priority.

Oliver heard that IBM pretty much ships these exact bits without additional tests.

• AlexA believes that they have additional confidence suites
  • Oliver wonders if can discuss with contributing companies to make a case for a bit of additional effort/resource to get the packaging done.
  • There was not general agreement that this request would be successful

Needs the details from dave williams to gauge what the true cost would be.

• It could be that WTP we can already know that the support cost would be high
• We hope not.

Weekly schedules

Oliver notes that trace and monitoring lists nothing but test passes in weekly schedule right now. Is this correct?

• Yes.
• Want to complete I5 iteration this week then NEXT week work on new defects.
• Has started assinging defects to next week.

Test

• W number of things have been deferred/slipped.
• They were a backlog for a while. looks worse than it is.
• Oliver asks if this is a problem for release.
  • Paul doesn't see an issue at this time

AlexN asks if move a bug do we account it as slipped from its original date in I5 to the position in I6?

• Some defects will fall completely out of 4.5. At that time stop tracking the slip for those bugs.
• There should be no slips for next week

Platform

• Joanna notes they have about 200 bugs to fix and some testing to do.
• Joanna put details just for 4.5 not for 4.4.1.
  • Have fixed 8-10 severe defects in last week or so.
  • Are doing continuous testing (because of test pass)
  • In process of triaging remaining 4.5 defects.
    • Joanna is trying to balance the load among team.
    • Will send out list to team in near future.
• Oliver asks whether all 200 bugs will get fixed by GA.
  • Joanna suggests that they probably will not.
    • Will bring forth a list for deferral about this time next week.

Oliver asks whether weekly schedule is providing any value to Joanna (following up on discussion from a few weeks ago)

• Right now it is more retrospective instead of proactive for Joanna.
• She is working to make it more proactive to try to make it more useful to her

Oliver notices that there is no weekly status and no bugs in trace

• Are no bugs to fix or no work planned?
  • AlexA reiterates that All bugs in profiler are managed under platform
  • Currently there are no bugs being efforted under trace.

Oliver asks again: there is a project called Trace. Are there any items there targeted for 4.5. What is Intel doing?

• AlexA notes that pretty much all that Intel is doing is managed under platform
• Paul noted that project plan has 4-5 defects for trace in btm area.
  • AlexA is tracking these things
  • Wants to discuss with matt and richard more detail to place them on schedule
  • Has a meeting scheduled to talk to them on Monday
  • Should be able to place on schedule at that time.
  • AI to AlexA to place the 4 trace items on schedule by next week's PMC.

Next week everyone should have I6 schedule taken care of.

Oliver asks if there are any other items related to 4.4.1 or 4.5?

• no

MISC

Team noted Monday is a holiday in North America

• US: presidents day
• Canada: Family day

Oliver on vacation next week but will try to make PMC call.

BVT update

• Paul updates PMC that they are running in a hacked mode.
  • AI Chris to send link for most recent report.
    • A report for the 2/11 build can be found here
  • They will be running routinely kicking off ~9-10am and finishing at about 11.
• Every day there will be a test report.
  • There will be a bvt report next to the build report link in the downloads page.

Paul and team will run this way for a week or so this way to work out the kinks then add AGR.

• Will schedule an AG type call to discuss it
• Explain how folks can add add tests to it (and remove them from their existing pass)

Oliver optimistically asks if any of his report stats change?

• Not yet because its too early.
• By about I6 test pass should start seeing such a measurement.

Paul mentioned an enhancement request from Asaf

• documentation for standalone packaging: bug 218464.
• AG review done. looks okay
• Officially approved by PMC

Chris raised discussion about AlexA as lead for Trace and issues with the portal.

• Guru said he didn't have appropriate rights to change over to AlexA
• Paul confirmed that for tptp.performance project the lead is set as Guru
• Guru may not have been able to get into portal correctly
  • Joanna had similar problem earlier on
• tptp.performance should be the pane that guru has access to
  • AI to Chris to try again with guru.
  • if not working then Oliver can send mail to resolve.