[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
|
[news.eclipse.tools.hyades] Meeting Notes: 13 - Feb - 2003
|
Execution Environment Control Group Members:
Kent Siefkes, Joe Toomey, Kris Kobylinski, Ian Hunter, Clay Williams, Neil
Sanderson (just joined the group this meeting)
Attending this meeting:
Kent Siefkes, Joe Toomey, Kris Kobylinski, Ian Hunter, Neil Sanderson
Proposed Agenda:
1. Pick weekly meeting time
2. Review current status against preliminary plan-for-plan we outlined at
the face-to-face meeting
3. Add any new tasks to the plan-for-plan.
4. Identify owners for remaining tasks, set target completion dates.
5. Identify potential white-paper(s) for the Execution group.
6. If time, discuss terminology and overlap with Data Collection group
7. Discuss committer(s) for the group
Actions/Results:
1.. Weekly meeting time was decided to be Thursday 10 - 11 a.m. EST, via
webex. (Contact Joe Toomey for a meeting invite if you're not in the
group, but would like to attend any of the meetings.)
2.. The preliminary plan-for-a-plan is copied below, and annotated for
progress, and further refinement of actions: (I've also attached Joe's notes
from the original breakout session, which is a good background for
testability interface and why we've coined the term "remote entities", which
is "everything to the right of the RAC" on the remote machine (since we had
a terminology impasse J):
1) Kris will provide Interface Descriptions from both workbench and
Remote Entities
1. Kris has sent general info on RAC and XMI fragments for agent
communication to hyades-team - DONE
2. Kris will prepare sequence diagram outlining use cases from existing
manual test execution environment as starting point for discussion (Kris
will mail out and present at next week's meeting)
3. Richard Duggan not ready yet to release complete interface
descriptions(1-2 weeks out yet); in the interim, Kris will prepare an
outline of the APIs from existing implementation by next week's meeting
2) Team identifies requirements for Deployment Remote Entity.
1. No progress yet on this, but we agree that we need this, at minimum
to deploy test assets necessary to run the test (even if we can leverage an
Eclipse plug-in to deploy SUT to app-server, etc.)
2. Not common agreement yet on just how this should be architected or
even its scope, but Kris has suggested 3 different aspects we will discuss
further.
3) Team identifies requirements (functional and non-functional) for
Remote Entities for Hyades
No progress yet beyond breakout meeting, but next steps were refined to be:
1. Define for common remote entity
2. Identify specializations of remote entities (Logging/Tracing,
Execution/Interaction, Deployment, .)
3. Define for specializations appropriate for this group
(Execution/Interaction, Deployment, ?)
4) Kris will present existing Remote Entities.
Limited currently to Logging Agents and Profiling Agents (defer until after
#3)
5) Determine work items, assign ownership, assign target dates
This is ongoing, we've added some work items in this meeting, in addition to
those above (see below).
3.. Additional work items were identified and discussed, including
volunteers for most of them:
1. Testability Interface (Ian overall)
i. Education,
examples on use of testability interfaces in scapatech's current product
(Ian)
ii. Define WDSL ->
EMF requirements (Ian)
iii.
Sample/reference implementation examples, using WDSL (Ian)
iv. Whitepaper on
how to build wrappers (using WDSL) - TBD
v. Plan for
developing "universal TRI connector" - TBD
2. Terminology - team
i. Identify and
describe execution-related elements needing names
ii. propose names
iii. get buy-in
from larger Hyades group
3. Execution Interop (or asset re-usability from existing frameworks)
i. Stax / Staf
(IBM open source)
1. education/examples (Clay/Joe?)
2. proposal for interop (Clay/Joe ?)
ii. TSS (Rational,
part of Test Manager)
1. education/examples (Kent)
2. proposal for interop (Kent)
iii. executing test scripts from other test tools - Kris, Kent
4. Identify requirements for API and/or architecture/model requirements
to facilitate inter/intra-test (virtual user) communication,
synchronization, shared services (shared variables, datapools, pooled
transaction rates, etc.) - Neil, Kent
i. Intracomputer
(on same "agent machine")
ii. Intercomputer
(on different "agent machines", participating in same test suite)
5. Identify API requirements for our Remote entities necessary to
execute the (new) test model - Kris
6. Any additional requirements imposed on datacomm interface - team
These are beyond the 4 requirements identified in the breakout session.
7. Educate each other about our respective execution environments in
order to garner a broader understanding of the range of requirements on
Hyades in terms of execution:
i. Scapatech -
Neil
ii. TestManager -
Kent
4.. See owners to tasks above. Target completion dates will follow.
5.. The only white paper identified so far is the one identified above on
how to build wrappers (using WDSL) for the testability interface
6.. Didn't have time to discuss terminology and overlap with Data
Collection group
7.. We discussed committers for the group, but were not sure how a single
committer for the team would work, per Harm's proposal. Within this group
we identified a committer for each company: Rational - Joe, Scapatech - Ian,
IBM - Kris. If we need a single lead committer, it would be Joe.