opsawg meeting IETF 86, Orlando, FL Tuesday, March 12, 1p-3p Notetaker: David Black, EMC --------------------------- 1. Benoit Claise (AD) - Discussion on OPSAWG document quality. Suggestions: 1. reach out to other people for review, 2. potentially put the bar higher for WG adoption (like in BMWG) 3. solicit reviews from chairs and AD 4. if no feedback, kill it. Demonstrate support. 5. bring the expert to the WG, e.g., as guest speaker. 2. LAG/ECMP Link Utilization (draft-krishnan-opsawg-large-flow-load-balancing), Ram Krishnan Curtis Villamizar: Looks like implementation details, not operational issues. Not appropriate here, even as an Informational RFC. Also has concerns with specific approach, which requires configuration. Ram Krishnan: Operational considerations for managing load balancing of large flows. Scott Bradner (WG co-chair): Need to think about Curtis's concern - if he is correct and there's no effect on "bits on the wire", this draft is likely to be out of scope. Melinda Shore (WG co-chair): Please make interoperability aspects of this clear. Chris Liljenstolpe (WG co-chair): And operational aspects. 3. CGN Deployment with BGP/MPLS IP VPNs (draft-ietf-opsawg-lsn-deployment) Victor Kuarsingh Lots of small changes, see slides. Looks like multicast considerations should be added in next version. -03 version coming, would like WG Last Call on that version. Scott Bradner (WG co-chair): Draft authors need to go seek out reviewers to take a hard look. Want to see feedback on list. Process request: Put *SLIDE NUMBERS* on slides so that remote participants can follow. 4. Firewalls (draft-ietf-opsawg-firewalls) Paul Hoffman There has been significant discussion on list. Proposed change in document focus: Plan to take out recommendations for firewall operators and vendors. Recommending taking out unused or worse features is unlikely to have any practical impact. Focus on operational effects of things firewalls could do without making recommendations on whether firewalls should do those things. Avoids getting caught between operational and security considerations/goals. Refocus on descriptions of what firewalls do for a network operator audience. Overall focus is serious network firewalls, not home Internet gateways. Ram Krishnan: Expand scope to related security detection mechanisms. Chris Liljenstolpe (as individual): Consider distributed firewall functionality, e.g., for large data centers. Paul Hoffman: Definitely in scope, this is a "firewall" even though it has a distributed implementation. Russ Mundy: Explain differences between user interfaces and underlying functionality. Paul Hoffman: Yes, but w/limited depth. Joel Jaeggli (AD): Taxonomy of firewalls and functionality is helpful. Chris Liljenstolpe (WG co-chair): Might want to take up related mechanisms in a separate draft. Melinda Shore (WG co-chair): Don't want to water document down to have no value. WG chairs need to manage "firewalls are evil" and related discussions around this draft. 5. RFC 5066bis status (draft-ietf-opsawg-rfc5066bis) Dan Romascanu Carrying forward only the IF-CAP-STACK-MIB in IETF, IEEE is taking EFM-CU-MIB forward. Will not obsolete RFC 5066, but 5066bis will update it. 5066bis will reference existing IF-CAP-STACK-MIB module in RFC 5066 instead of copying it into the new 5066bis draft. Will write a document on transition of this work to IEEE, similar to RFC 4633 on the Bridge MIB. 6. CAPWAP Related drafts, Hui Deng draft-chen-opsawg-capwap-extension draft-shao-opsawg-capwap-hybridmac draft-zhang-opsawg-capwap-eap Need extensions to CAPWAP, but CAPWAP WG no longer active. Have broken material into 3 drafts, 2 for standards track, 1 for informational. 6.1 First draft: Hybrid MAC model. Model will be informational. 6.2 Second draft: 802.11n extensions, Dapeng Liu is presenter Comments: This draft is very important and useful. 6.3 Third Draft : EAP encap for CAPWAP control plane, Rong Zhang is presenter 6.4 Discussion: Do what with these drafts? Implementations cover both drafts. Joel Jaeggli (AD): Have requested external review of these docs by CAPWAP experts. Dan Romascanu: Should also solicit reviews from IEEE 802.11. Dan will facilitate with 802.11 liaison to obtain IEEE 802.11 expert review(s). That should happen before opsawg adopts these as WG drafts. Scott Bradner (WG co-chair): Need reviews, revise docs according to results of reviews, then can ask for adoption. Sense of room is that documents should become WG documents. Adoption will be handled on list. 7. Deterministic CGN draft, Chris Donley draft-donley-behave-deterministic-cgn-05 Request for detailed review of this draft. Obtained commitments to 3 reviews in next couple of weeks. 8. Virtual Machine MIB (draft-asai-vmm-mib), Hirochika Asai Have merged two prior drafts into one, use general UUID TC, add read/write I/O statistics. Discussion of VM states and state machine - does that belong in a MIB document? Could be a reasonable location, as operational concerns force agreement on a common abstract model, or move into a separate document. This material is likely to be useful to nvo3 WG. Request for more reviews.