Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa13741; 15 Apr 93 15:06 EDT Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa13737; 15 Apr 93 15:06 EDT Received: from interlock.ans.net by CNRI.Reston.VA.US id aa26516; 15 Apr 93 15:06 EDT Received: from lager.cisco.com by interlock.ans.net with SMTP id AA20171 (InterLock SMTP Gateway 1.1 for ); Thu, 15 Apr 1993 14:54:55 -0400 Received: by lager.cisco.com; Thu, 15 Apr 1993 11:55:49 -0700 Date: Thu, 15 Apr 1993 11:55:49 -0700 Sender:ietf-archive-request@IETF.CNRI.Reston.VA.US From: Tony Li Message-Id: <199304151855.AA28132@lager.cisco.com> To: iwg@ans.net Subject: New draft Hi all, A new draft of BGP 4 containing all of the changes from Columbus is now in ftp.cisco.com:tli/bgp.txt. If there are no comments for two weeks, I will submit this as an Internet Draft. If there are no comments to that for two weeks, we will move it to being an RFC. Tony   Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa13860; 15 Apr 93 15:08 EDT Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa13856; 15 Apr 93 15:08 EDT Received: from interlock.ans.net by CNRI.Reston.VA.US id aa26599; 15 Apr 93 15:08 EDT Received: from watson.ibm.com by interlock.ans.net with SMTP id AA20213 (InterLock SMTP Gateway 1.1 for ); Thu, 15 Apr 1993 14:59:53 -0400 Message-Id: <199304151859.AA20213@interlock.ans.net> Received: from YKTVMV by watson.ibm.com (IBM VM SMTP V2R3) with BSMTP id 7163; Thu, 15 Apr 93 15:00:47 EDT Date: Thu, 15 Apr 93 14:59:39 EDT Sender:ietf-archive-request@IETF.CNRI.Reston.VA.US From: yakov@watson.ibm.com To: tli@cisco.com, iwg@ans.net Subject: New draft Ref: Your note of Thu, 15 Apr 1993 11:55:49 -0700 >we will move it to become an RFC... Small correction: the WG will recommend to advance it to a Proposed Standard. Yakov   Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa20162; 19 Apr 93 12:35 EDT Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa20157; 19 Apr 93 12:35 EDT Received: from interlock.ans.net by CNRI.Reston.VA.US id aa26191; 19 Apr 93 12:35 EDT Received: from moe.rice.edu by interlock.ans.net with SMTP id AA30639 (InterLock SMTP Gateway 1.1 for ); Mon, 19 Apr 1993 12:02:09 -0400 Received: from watson.ibm.com by moe.rice.edu (AA19618); Mon, 19 Apr 93 11:03:03 CDT Message-Id: <9304191603.AA19618@moe.rice.edu> Received: from YKTVMV by watson.ibm.com (IBM VM SMTP V2R3) with BSMTP id 2339; Mon, 19 Apr 93 12:03:03 EDT Date: Mon, 19 Apr 93 11:58:37 EDT Sender:ietf-archive-request@IETF.CNRI.Reston.VA.US From: yakov@watson.ibm.com To: iwg@rice.edu Subject: mailing list updates. Folks, Just to let you know that the *real* mailing list for BGP WG is maitained by ANS: iwg@ans.net or bgp@ans.net. Please try to avoid iwg@rice.edu, so that we'll be able to migrate to the list maintained by ANS and retire iwg@rice.edu. Thanks. Yakov P.S. I would like to take this opportunity to thank ANS for the willingness to maintain the mailing list.   Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa09175; 21 Apr 93 12:35 EDT Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa09171; 21 Apr 93 12:35 EDT Received: from interlock.ans.net by CNRI.Reston.VA.US id aa18595; 21 Apr 93 12:35 EDT Received: from moe.rice.edu by interlock.ans.net with SMTP id AA11533 (InterLock SMTP Gateway 1.1 for ); Wed, 21 Apr 1993 12:11:04 -0400 Received: from watson.ibm.com by moe.rice.edu (AA18391); Wed, 21 Apr 93 11:11:58 CDT Message-Id: <9304211611.AA18391@moe.rice.edu> Received: from YKTVMV by watson.ibm.com (IBM VM SMTP V2R3) with BSMTP id 8671; Wed, 21 Apr 93 12:11:55 EDT Date: Wed, 21 Apr 93 12:11:14 EDT Sender:ietf-archive-request@IETF.CNRI.Reston.VA.US From: yakov@watson.ibm.com To: iwg@rice.edu Subject: proposed changes to the BGP WG mailing list Folks, As you may know the actual BGP WG mailing list is maintained at ANS (bgp@ans.net). Therefore, I would suggest to dismantle the old mailing list (iwg@rice.edu) and ask folks at Rice to send back a canned message informing the sender of the change and NOT forwarding the mail to bgp@ans.net. If nobody in the list will object to this suggestion by 5/31/93, we'll implement the proposed change by 6/1/93. Yakov.   Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa10389; 21 Apr 93 14:02 EDT Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa10385; 21 Apr 93 14:02 EDT Received: from interlock.ans.net by CNRI.Reston.VA.US id aa21822; 21 Apr 93 14:02 EDT Received: from moe.rice.edu by interlock.ans.net with SMTP id AA11745 (InterLock SMTP Gateway 1.1 for ); Wed, 21 Apr 1993 13:38:34 -0400 Received: from is.rice.edu (rg-90.rice.edu) by moe.rice.edu (AA19741); Wed, 21 Apr 93 12:39:28 CDT Received: from sabine.is.rice.edu by is.rice.edu (AA05556); Wed, 21 Apr 93 12:39:25 CDT Received: by sabine.is.rice.edu (AA06984); Wed, 21 Apr 93 12:39:25 CDT Sender:ietf-archive-request@IETF.CNRI.Reston.VA.US From: William Manning Message-Id: <9304211739.AA06984@sabine.is.rice.edu> Subject: Re: proposed changes to the BGP WG mailing list To: yakov@watson.ibm.com Date: Wed, 21 Apr 93 12:39:24 CDT Cc: iwg@rice.edu In-Reply-To: <9304211611.AA18391@moe.rice.edu>; from "yakov@watson.ibm.com" at Apr 21, 93 12:11 pm X-Mailer: ELM [version 2.3 PL11] Sounds good to me. -- Regards, Bill Manning bmanning@rice.edu PO Box 1892 713-285-5415 713-527-6099 Houston, Texas R.U. (o-kome) 77251-1892   Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa14277; 29 Apr 93 17:35 EDT Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa14273; 29 Apr 93 17:35 EDT Received: from interlock.ans.net by CNRI.Reston.VA.US id aa01210; 29 Apr 93 17:35 EDT Received: from moe.rice.edu by interlock.ans.net with SMTP id AA04898 (InterLock SMTP Gateway 1.1 for ); Thu, 29 Apr 1993 17:06:03 -0400 Received: from is.rice.edu (rg-90.rice.edu) by moe.rice.edu (AA26193); Thu, 29 Apr 93 16:06:58 CDT Received: from sabine.is.rice.edu by is.rice.edu (AA09423); Thu, 29 Apr 93 16:06:57 CDT Received: by sabine.is.rice.edu (AA00502); Thu, 29 Apr 93 16:06:55 CDT Sender:ietf-archive-request@IETF.CNRI.Reston.VA.US From: William Manning Message-Id: <9304292106.AA00502@sabine.is.rice.edu> Subject: proposed changes to the BGP WG mailing list To: iwg@rice.edu Date: Thu, 29 Apr 93 16:06:54 CDT X-Mailer: ELM [version 2.3 PL11] > > Folks, > As you may know the actual BGP WG mailing list is maintained > at ANS (bgp@ans.net). Therefore, I would suggest to dismantle > the old mailing list (iwg@rice.edu) and ask folks at Rice to > send back a canned message informing the sender of the change > and NOT forwarding the mail to bgp@ans.net. > > If nobody in the list will object to this suggestion by 5/31/93, > we'll implement the proposed change by 6/1/93. > > Yakov. > Per this recommendation, the iwg@rice.edu mailing list is now disabled. Please verify that you are on the correct mailing list by sending a request to: bgp-request@ans.net The actual implementation of a canned message will occur on the scheduled date. Thanks. -- Regards, Bill Manning bmanning@rice.edu PO Box 1892 713-285-5415 713-527-6099 Houston, Texas R.U. (o-kome) 77251-1892   Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa01035; 22 Apr 93 13:40 EDT Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa01030; 22 Apr 93 13:40 EDT Received: from interlock.ans.net by CNRI.Reston.VA.US id aa03797; 22 Apr 93 13:40 EDT Received: from mitchell.cit.cornell.edu (NR-TECH.CIT.CORNELL.EDU) by interlock.ans.net with SMTP id AA28216 (InterLock SMTP Gateway 1.1 for ); Thu, 22 Apr 1993 12:04:41 -0400 Received: from MITCHELL.CIT.CORNELL.EDU by mitchell.cit.cornell.edu (4.1/1.34/Honig-1.3) id AA04620; Thu, 22 Apr 93 12:05:36 EDT Message-Id: <9304221605.AA04620@mitchell.cit.cornell.edu> To: bgp@ans.net Subject: Error in draft Date: Thu, 22 Apr 1993 12:05:35 -0400 Sender:ietf-archive-request@IETF.CNRI.Reston.VA.US From: David J LeRoy I hope it's not to late for input regarding the latest bgp draft. I just wanted to point out that in Appendix A concerning the FSM, in the table that summarizes the state transitions, the Active and Connect column labels are switched. The actual contents of each column are not switched just the labels. The table should read: Events| Idle | Connect| Active | OpenSent | OpenConfirm | Estab | (1) | (2) | (3) | (4) | (5) | (6) |-------------------------------------------------------------- 1 | 2 | 2 | 3 | 4 | 5 | 6 | | | | | | 2 | 1 | 1 | 1 | 1 | 1 | 1 | | | | | | 3 | 1 | 4 | 4 | 1 | 1 | 1 | | | | | | 4 | 1 | 1 | 1 | 3 | 1 | 1 | | | | | | 5 | 1 | 3 | 3 | 1 | 1 | 1 | | | | | | 6 | 1 | 1 | 1 | 1 | 1 | 1 | | | | | | 7 | 1 | 2 | 2 | 1 | 1 | 1 | | | | | | 8 | 1 | 1 | 1 | 1 | 1 | 1 | | | | | | 9 | 1 | 1 | 1 | 1 | 5 | 6 | | | | | | 10 | 1 | 1 | 1 | 1 or 5 | 1 | 1 | | | | | | 11 | 1 | 1 | 1 | 1 | 6 | 6 | | | | | | 12 | 1 | 1 | 1 | 1 | 1 | 1 or 6 Regards, Dave LeRoy dleroy@mitchell.cit.cornell.edu (919)-254-9136   Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa09377; 23 Apr 93 14:42 EDT Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa09373; 23 Apr 93 14:42 EDT Received: from interlock.ans.net by CNRI.Reston.VA.US id aa22364; 23 Apr 93 14:41 EDT Received: from watson.ibm.com by interlock.ans.net with SMTP id AA10635 (InterLock SMTP Gateway 1.1 for ); Fri, 23 Apr 1993 14:22:43 -0400 Message-Id: <199304231822.AA10635@interlock.ans.net> Received: from YKTVMV by watson.ibm.com (IBM VM SMTP V2R3) with BSMTP id 3197; Fri, 23 Apr 93 13:14:01 EDT Date: Fri, 23 Apr 93 13:12:41 EDT Sender:ietf-archive-request@IETF.CNRI.Reston.VA.US From: yakov@watson.ibm.com To: bgp@ans.net Subject: minutes of the WG meeting Folks, Appended are the minutes of the joint BGP/IDRP for IP WGs meeting in Columbus. Please review them asap. Yakov. --------------------------------------------------------------------- Minutes of the Jointt IDRP/BGP WG meeting; Columbus IETF -------------------------------------------------------- Minutes were taken by John Scudder (MERIT) and Susan Hares (MERIT) PIP routing issues ------------------ - Short overview by Bala Rajagopalan of PIP (get the slides or PIP docs if you want details) - Concerns about functionality the PIP people think is missing from IDRP: Concern: Ability to meaningfully carry PIP addresses. Response: IDRP can do this just fine, since it views the NLRI as an opaque blob. Concern: Ability to represent multiple addressing levels (so- called "clusters") Response: Looks like this maps nicely into RDCs. Note that it is a non-requirement of PIP to be able to heal a partitioned cluster via connections external to the cluster. Concern: Autoconfiguration of BIS neighbors. Response: None. (correction (Sue Hares) - see notes on IS-IS to IDRP from ANSI to give ideas on this) Concern: Configuration of adjacencies with distant BISes via tunneling. Response: This is in the works. Concern: Lack of (PIP-style) QoS support. Response: Use locally-defined QoS. Note that locally-defined QoS need not map in any way to 8473. - It was asked whether MPLV is required for PIP routing. Answer was no, path vector is desired but MPLV in specific is not required. - Necessary actions to integrate PIP into IDRP: 1. assign PIP ID 2. assign RD space for PIP 3. Further work on tunneling IDRP for IP issues: ------------------- - The following documents have been completed and will be submitted: 1. IDRP documents overview 2. IDRP family document 3. MIB document - No application document has been written, nor is one planned. It was suggested that the BGP/CIDR documents and ISO 10747 provide the equivalent of an application document. Sue will add references to these documents into the family document. - It was observed that there exists no document discussing the overall architecture of the Internet. This is outside the scope of the BGP and IDRP groups and will be referred to Hinden et al. - Agreement to move IDRP for IP to proposed standard as soon as working implementations are available. - Standardization update: The official ballot closes at the end of April; the DIS is expected to move to full standard at that time. BGP-4 issues: ------------- - Hinden requested that BGP-4 be advanced to proposed standard. - One implementation is currently available (BGP-4 in gated implemented by Dennis Fergusen). - The following documents have been completed: 1. MIB 2. Usage document - The protocol spec requires some editing (see below); after editing it will be issued as an Internet Draft and then progressed to proposed standard. Changes to BGP-4 spec: ---------------------- - editorial changes 1. sense of LOCAL_PREF has been reversed: Higher value is now preferred. Reason: Self-consistency + Consistency with BGP-3. There was some opinion that this was a bad idea. 2. Grammar fixes 3. Clarification that mandatory attributes do not have to occur if the message carries no NLRI. - Size of LOCAL_PREF and MULTI_EXIT_DISC 1. The sizes of the LOCAL_PREF AND MULTI_EXIT_DISC attributes will be changed from 8 bits to 32 bits. - Other changes: 1. An IP protocol ID still needs to be assigned for IDRP. (45 has been assigned by the IANA) 2. An IS-IS/IDRP interaction document was requested. Sue responded that the theory is already covered by ANSI documents X3S3.3/92-162R). However, this document only addresses IS-IS for CLNP, not for IP, so an IS-IS for IP/IDRP document is still needed. Stability on the NSFNET backbone on the Network: (Curtis Villamizar) ----------------------------------------------- 1.) collection of the notes on the Stability of Internal BGP Peers: factored out of this: long circuit outages not factored out of this: configuration runs single link drops monthly summary of the IBGP connection up time. - 99% of time all ibgp session were up for all routers. (1% of remaining time is majorily impacted by configuration runs which take down the IBGP sessions.) Last October, and the last IETF in November we had days where at least one router was flapping at least 20-30% of the time. The problem which caused these flaps were inefficenct code redistribing of EGP learned routes into the IBGP. One IBGP packet was sent per route rather than aggregating of the routes. This packing of routes (one/pdu) caused 1000 times more calls to the operating system than necessary. That plus an extremely high external route flapping from other networks caused ANSNET stability problems. External Stability: 90-100% uptime for external routers. Large spikes in external uptime were caused 3 isolated events concern a single flapping of peer router for long time period. External Route flap: Periods of flapping include aproximate the 1000 - 2500 routes per hour. Difference between short periods of time to 15 minutes samples is that a great deal of routing flaps are missed. A high percentage of route flapping occurs at 2-6 minutes time period. A Damping algorithm proposed by Curtis is that a minimum timer for announcing timers be calculated by the timers based on the stability of the route. Such as each time it flaps, the timer value for holding down the connection is re-calculated. Unknown how much of flapping would be covered by aggregation. OSPF External Attributes Overview (Dennis Fergusson) --------------------------------- Document on this is: draft-ietf-osp-ext-attr-00.txt Dennis Ferguson gave an overview the changes he's proposing to OSPF to carry external attributes in the OSPF. For BGP, this means the AS path information and the origin. The benefit of this approach is that External attributes would be flooded only once. A snapshot on the Sunday before IETF, found an average of 30 networks per distinct set of attributes, with about 1.3 bytes per network overhead for the OSPF LSA additions. As part of the overview the following issues were discussed: 1.) Local preference is a useful mechanism for the passing of the policy. With the proposed changes to OSPF, OSPF metric does sort of the same thing and the local preference wouldn't be necessary. 2.) What will OSPF WG do on this? It was suggested that OSPF WG will publish it as a simple extension to OSPF. 4.) What happens if a OSPF router does not propagate this? Need a propagating path between border routers to implement this. Dennis published a paper to describing the propagation. Only know problem is that the there is a known start-up problem for OSPF where the router could get confused about the routers. The problem can be avoided by not configuring network in such a way, that it is sparsely connected for type 8 connections. Randomize the mechanism the selecting router id to minimize the connections interlock. 5.) How would this work: Mapping of OSPF TOS 7 bits to IDRP bits. 5 pages on this document. BGP4/IDRPfor IP -- OSPF Interaction Document (Kannan Varadhan) --------------------------------------------- Current mapping between IDRP/BGP4 and IDRP BDGP/IDRP MULTI_EXIT_DISC - BGP ORIGIN and AS_PATH OSPF cost and type, IDRP EXT_INFO and RD_PATH OSPF tag BGP/IDRP NEXT_HOP OSPF Forwarding Address BGP/IDRP LOCAL_PREF OSPF cost Minimal Implementation of the specification, set of sufficient clauses that ensure the safety of the Internet. 1) Must not advertise a route containing a set of reachable destination when none the destination in the address/ mask pair is reachable via OSPF (section 2.1, bullet 3) 2) Must advertise an externally received route into OSPF if this route is selected at the completion of the phase 2 route selection process. (section 2.2, bullet 1) 3) Must set the BGP/IDRP dentifier to be the same as the OSPF router ID. (section 3) 4) MUST set the OSPF tag accurately (section 4) Section 2.2) Importing BGP/IDRP information into OSPF Dennis Fergusson did not like the section on BGP speaker. He will provide new text.   Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa16437; 26 Apr 93 12:12 EDT Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa16433; 26 Apr 93 12:12 EDT Received: from interlock.ans.net by CNRI.Reston.VA.US id aa24757; 26 Apr 93 12:12 EDT Received: from lager.cisco.com by interlock.ans.net with SMTP id AA06110 (InterLock SMTP Gateway 1.1 for ); Mon, 26 Apr 1993 11:49:19 -0400 Received: by lager.cisco.com; Sat, 24 Apr 1993 23:26:40 -0700 Date: Sat, 24 Apr 1993 23:26:40 -0700 Sender:ietf-archive-request@IETF.CNRI.Reston.VA.US From: Tony Li Message-Id: <199304250626.AA03913@lager.cisco.com> To: dleroy@nr-tech.cit.cornell.edu Cc: bgp@ans.net In-Reply-To: David J LeRoy's message of Thu, 22 Apr 1993 12:05:35 -0400 <9304221605.AA04620@mitchell.cit.cornell.edu> Subject: Error in draft David, Thanks, it's never too late. I've fixed the master copy and ftp.cisco.com:tli/bgp.txt. Tony   Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa14968; 29 Apr 93 18:01 EDT Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa14964; 29 Apr 93 18:01 EDT Received: from interlock.ans.net by CNRI.Reston.VA.US id aa02252; 29 Apr 93 18:01 EDT Received: by interlock.ans.net id AA20809 (InterLock SMTP Gateway 1.1 for bgp@ans.net); Thu, 29 Apr 1993 17:46:51 -0400 Received: by interlock.ans.net (Internal Mail Agent-2); Thu, 29 Apr 1993 17:46:51 -0400 Received: by interlock.ans.net (Internal Mail Agent-1); Thu, 29 Apr 1993 17:46:51 -0400 Date: Thu, 29 Apr 93 17:47:40 EDT Sender:ietf-archive-request@IETF.CNRI.Reston.VA.US From: David Bolen To: bgp@ans.net Subject: Re: proposed changes to the BGP WG mailing list Message-Id: > Date: Thu, 29 Apr 93 16:06:54 CDT > From: bmanning@is.rice.edu (William Manning) > Subject: proposed changes to the BGP WG mailing list > > Please verify that you are on the correct mailing list by sending a > request to: > > bgp-request@ans.net So that our postmaster doesn't get bogged down with a number of such requests, please don't bother sending such a note. The current list "iwg@rice.edu" simply forwards the request on to "iwg@ans.net", which is synonymous with "bgp@ans.net". The actual list membership has been maintained here at ANS for some months already. If you received Bill's note then you are definitely on the "bgp@ans.net" list, since that is how Bill's note got to you in the first place. Requesting verification of this is really superfluous. Thanks. -- David /-----------------------------------------------------------------------\ \ David Bolen \ Internet: db3l@ans.net / | Advanced Network & Services, Inc. \ Phone: (914) 789-5327 | / 100 Clearbrook Road, Elmsford, NY 10523 \ Fax: (914) 789-5310 \ \-----------------------------------------------------------------------/   Received: from ietf.nri.reston.va.us by IETF.CNRI.Reston.VA.US id aa15910; 29 Apr 93 18:55 EDT Received: from CNRI.RESTON.VA.US by IETF.CNRI.Reston.VA.US id aa15906; 29 Apr 93 18:55 EDT Received: from interlock.ans.net by CNRI.Reston.VA.US id aa03573; 29 Apr 93 18:55 EDT Received: from is.rice.edu by interlock.ans.net with SMTP id AA07484 (InterLock SMTP Gateway 1.1 for ); Thu, 29 Apr 1993 18:32:03 -0400 Received: from sabine.is.rice.edu by is.rice.edu (AA10827); Thu, 29 Apr 93 17:32:58 CDT Received: by sabine.is.rice.edu (AA00550); Thu, 29 Apr 93 17:32:12 CDT Sender:ietf-archive-request@IETF.CNRI.Reston.VA.US From: William Manning Message-Id: <9304292232.AA00550@sabine.is.rice.edu> Subject: Re: proposed changes to the BGP WG mailing list To: David Bolen Date: Thu, 29 Apr 93 17:32:12 CDT Cc: bgp@ans.net In-Reply-To: ; from "David Bolen" at Apr 29, 93 5:47 pm X-Mailer: ELM [version 2.3 PL11] Not quite true Dave. I have removed the pointer to the BGP@ANS.NET list. This mail went only to those on the original iwg@rice.edu list. I have no idea how many folks are not on both. This is an attempt to verify that no one gets dropped intentionally. -- Regards, Bill Manning bmanning@rice.edu PO Box 1892 713-285-5415 713-527-6099 Houston, Texas R.U. (o-kome) 77251-1892