Jump to: navigation, search

Difference between revisions of "Meetings/VPNaaS"

(Bucket List)
(Meetings)
Line 5: Line 5:
 
* Chair: pc_m (Paul Michali)
 
* Chair: pc_m (Paul Michali)
  
Well, this is embarrassing...
+
So we have three votes for option (C), different day/time and on-demand, with times of 1600 UTC or later, 1630 UTC, and 1600 UTC on Tuesday. We have one vote for option (A), same day and time (Tuesday 1500), but on-demand. One vote for (D) include in Neutron meeting as on-demand topic (for now).
  
We haven't been having much turn out for the meetings over the past several weeks. As a result, I sent a questionnaire out on February 24th to see if how we wanted to move forward with meetings. Well, there was an issue with my return address and the message never posted to the openstack-dev mailing list (and I didn't get a bounced mail response).  Thinking there was no interest (sorry), I took the default action and gave VPN status as part of the March 2nd Neutron meeting.
+
Since the consensus is on-demand, we'll need to decide how to determine when we want to hold the meetings (how to signal). It looks like the majority is 1600 UT, so unless there is a change, we can try to schedule future meetings using a different time.
  
Mathieu mentioned that he didn't see the email, so I re-sent the survey out. We'll have a meeting today, at the normal time and can discuss. Also, I'll give another week to see what people think and we can decide how to handle meetings forward going.
+
Let's discuss at the next meeting.
  
 
= Logs and Minutes=
 
= Logs and Minutes=

Revision as of 19:41, 9 March 2015

Meetings

  • Weekly on Tuesdays at 1500 UTC
  • IRC channel: #openstack-meeting-4
  • Chair: pc_m (Paul Michali)

So we have three votes for option (C), different day/time and on-demand, with times of 1600 UTC or later, 1630 UTC, and 1600 UTC on Tuesday. We have one vote for option (A), same day and time (Tuesday 1500), but on-demand. One vote for (D) include in Neutron meeting as on-demand topic (for now).

Since the consensus is on-demand, we'll need to decide how to determine when we want to hold the meetings (how to signal). It looks like the majority is 1600 UT, so unless there is a change, we can try to schedule future meetings using a different time.

Let's discuss at the next meeting.

Logs and Minutes

Meetings, with their notes and logs, will be found under http://eavesdrop.openstack.org/meetings/vpnaas/

Agenda

Updated March 3rd, 2015

Announcements

  • DevStack support for StrongSwan is now upstreamed.
  • Directory structure change for functional test area is upstreamed.
  • VPN agent refactoring has started - goal is to eliminate dependency on L3 agent.
  • New functional job being created for StrongSwan (out for review) - will run in experimental mode.
  • StrongSwan implementation is out for review - will target Ubuntu initially.
  • Scenario test out for review. Review 159746

Functional Testing

Plan is to have tests/functional/ hold shared functional test modules, tests/functional/openswan/ to hold OpenSwan specific tests, and tests/functional/strongswan/ to hold StrongSwan specific tests.

A new check-neutron-vpnaas-dsvm-functional-sswan is being created, to set up DevStack for StrongSwan operation.

The post-test-gate.sh script needs to be modified for support of scenario testing (159746).

Bugs

Current bugs: VPNaaS bugs

We need to get StrongSwan commit tested out and reviewed, so please help out (https://review.openstack.org/144391). This commit will now only address Ubuntu, and a separate commit (later) will address Fedora support (as there are some differences).

Bring up any other bugs of interest.

Bucket List

Here are some ideas for tasks that need to be done (feel free to work on them - put your name by any you choose)...

  • The 'docs' target needs to be implemented for our repo (doc file was removed from Neutron).
  • The old netconn-api documentation needs to be migrated to WADL so can be included into Reference API Wiki.
  • Coverage, especially in database modules, is lacking.
  • Need functional tests for OpenSwan device driver
  • Refactor duplication out of device driver code (OpenSwan, StrongSwan, Cisco, Vyatta) (Sridar R working on?)
  • The OpenSwan class should be separated from the ABC definition, and placed into a new module. (Brian Bowen working on)
  • Remove /n from execute method in utils.py so that duplicate code can be removed in VPN drivers.

Open Discussion

Charter

VPNaaS Team Charter

Meeting Commands

/join #openstack-meeting-4
#startmeeting vpnaas
#topic Announcements
#undo

...

#endmeeting