#opendaylight-group-policy: gbp_status_arch

Meeting summary

1. agenda (tbachman, 15:00:11)
   1. https://meetings.opendaylight.org/opendaylight-group-policy/2015/gbp_status_arch/opendaylight-group-policy-gbp_status_arch.2015-07-29-15.02.html Minutes from last week’s meeting (tbachman, 15:00:19)
   2. committers please #info in (tbachman, 15:00:24)
   3. ACTION: tbachman to talk with moizer and pantelis about colocation of shard (tbachman, 15:01:04)
   4. alagalah says one of the outcomes of the summit last week was discussions around pipeline coexistence and overlays (tbachman, 15:01:29)
   5. alagalah says the LISP mapper provides a lot of flexibility, so we’ll have follow-up conversations to understand their architecture (tbachman, 15:01:52)
   
   
2. ODL Summit (tbachman, 15:01:59)
   1. alagalah says we may be able to leverage the work in LISP flow mapping, but we need to make sure that there aren’t any corner cases (tbachman, 15:02:33)
   2. alagalah says he tested some ideas he had on pipeline coexistence — has spoken with ebrjohn about working on this in the SFCOFL2 renderer (ebrjohn says he’s fine with that) (tbachman, 15:03:33)
   3. tbachman says there was some discussion with the integration group on how to use/integrate docker (tbachman, 15:05:16)
   4. alagalah says we also had a meeting for our Beryllium planning release (tbachman, 15:05:29)
   5. alagalah says we have a strawman proposal for the plan, but is going to look into scheduling some meetings for more input — will provide this as the draft plan, but final isn’t due until M2 (tbachman, 15:06:27)
   6. alagalah says we’re also doing some CSIT work to help out the SFC project (since we’re doing the same in GBP) (tbachman, 15:07:02)
   7. alagalah says he proposed a state machine for SFC orchestration in the OPNFV-SFC call — will do the same on the ODL SFC call (tbachman, 15:07:58)
   8. alagalah says the key thing we’re focusing on now is stability — clustering, CSIT, etc. (tbachman, 15:08:17)
   9. alagalah also is looking into cleaning up the model — domain specifics and non-domain specifics (tenant model contains everything, and can/should be cleaned up) (tbachman, 15:11:57)
   10. alagalah says that instrumentation is another bucket of work that we need for beryllium (tbachman, 15:13:06)
   11. There are 3 parts in instrumentation; OAM, Statistics Gathering (state data, and time-series data) (tbachman, 15:13:38)
   12. alagalah says we’ll look at TSDR to see if there’s work there that can be leveraged (tbachman, 15:13:56)
   13. alagalah says there was also discussion about consumability for GBP — how do we make GBP more consumable (tbachman, 15:15:01)
   14. alagalah says that one solution is to investigate providing a simplified upper-layer API (tbachman, 15:16:47)

Action items

1. tbachman to talk with moizer and pantelis about colocation of shard

Action items, by person

1. tbachman
   1. tbachman to talk with moizer and pantelis about colocation of shard

People present (lines said)

1. tbachman (27)
2. odl_meetbot (4)
3. alagalah (0)