Operational Requirements Area Director(s): o Susan Estrada: estradas@cerf.net o Phill Gross: pgross@nis.ans.net o Bernhard Stockman: boss@sunet.se Area Summary reported by Bernhard Stockman/NORDUnet Operational Requirements Area Directorate (ORAD) The Operational Requirements Area Directorate has so far been an open Group with sessions at IETF meetings. The meeting discussed this time the formation of a closed Directorate similar to SAAG in the security area or USAC in the User Services Area. It was decided that ORAD will consist of 12-16 people with specific duties as described in the ORAD Job description. Meeting at IETF's will continue to be open. Two mailing lists will be created: one for ORAD as such and one, orad-interest@ans.net, for people interested in operational related issues. Job Description for ORAD. ORAD will be chartered to: o Approve Charters of proposed working groups. o Explicitly create necessary working groups. o Track progress of all Operational Requirements Area working groups. o Review *all* Internet Drafts for operational impacts. o Liaison with other IETF areas and external groups. o Others... Operational Concerns. A broad review was made over topics for ORAD. An extensive list was produced containing most operational related topics. For each topic a short description of the topic will be written down. Among the selected items some where regarded as more important and deserving immediate attention especially the routing and addressing problems. Current routing and addressing problems. This includes very short-term issues like reuse of IP address space and effects of more thrust in routers. An address assingment plan is under preparation as part of the CIDR concept. Proposed routing and addressing plans will will be reviewed by ORAD with respect to the operational impact such as: o Details of proposed schemes. 1 o Changes required in hosts and routers. o Performance impact. o Scaling properties. o Support for old host and routers. o Deployment plan. o Security impact. o Operational management and training impact. The IEPG initiated specification and implementation of a Global Internet eXchange (GIX) is part of the solution of the today routing problems and will accordingly be on the ORAD agenda. Other operational items mentioned were: o Globally coordinated DNS. o NIC/NOC coordination. o Network performance measurements. o Security. o OSI operations. IP Addressing Plan BOF (IPADDR) The intention with this BOF was to discussed the possibilities of an Internet Addressing plan as part of the CIDR deployment. Two Internet Drafts have been submitted proposing similar schemes. One main difference was the proposed size of blocks to be assigned. After the BOF it was decided to make the following recommendations: 1. Blocks are allocated initially to Europe and the US with (216.*.* - 219.*.* to Europe) and (220.*.* - 223.*.* to the US). 2. Blocks will be distributed at a size of (256*256 class C networks). 3. Class B address space will only be used in rare circumstances for example to international multihomed organizations. 4. Upper half of Class A address space will be reserved for future use. The BOF discussed issues related to the administration of delegation of authorities. It is recommended that there will be top level geographically based top authorities. BGP Deployment and Application Working Group (BGPDEPL) The Internet BGP topology was presented. There are about 25 administrative systems which are reachable from the NSFnet via EGP2, and 2 about 40 which are reachable via BGP. Peter Loethberg described the European EBONE BGP deployment. In several ways EBONE is further along in BGP deployment than US networks. Several vendors described the status of their implementations. CISCO and Cornell gated have BGP3 running today, and are working on BGP4. BBN has BGP3 running and is working on BGP4. Other vendors including Proteon and Wellfleet are actively working on BGP3 in preparation for BGP4. ANS intends to have BGP4 deployed January 1993 and has offered to help vendors with interoperability testing. Vendors can arrange to bring equipment into the ANS test facility. ANS is also exploring support for remote testing by ``tunneling'' BGP from the ANS test network through the NSFnet. Contact Jordan Becker for further information. Benchmarking Methodology Working Group (BMWG) Two proposed additional performance tests were discussed. The Group concluded that one of the proposals should not be adopted. The second proposal, to add application level performance test, should be discussed with the test equipment vendors to see if they would be willing to implement such tests. The latest draft was discussed and a few editorial changes were suggested. It was strongly urged that some additions be made to the draft to add the rational behind the various tests and procedures. Network Joint Management Working Group (NJM) The Operational Impact of the NSF recompetition was investigated. A mailing list - recompete@nsf.gov - has been created to discuss these topics. Papers are available at expres.cise.nsf.gov. Operational implications of address changes. How should N CIDR class C networks be routed internally. Application and Information Services was discussed with respect to troubleshooting and maintenance. There is a need for service level maps, e.g., video topology maps. The session continued with some networking ``war stories''. The BBN network 192.1.1 will be discontinued because some vendors of medical equipment use this network as their default! There are rumors that the SUN example network number was returned and is now re-issued to some unsuspecting site. Network Status Reports (NETSTAT) Four networks delivered reports: 3 1. EBONE - Bernhard Stockman 2. NASA Science Internet - Jeffry Burgan 3. ANS/Merit T1/T3 J- Jordan Becker/Elise Gerich 4. Brazil networking - Pushpendra Motha EBONE - KTH (The Royal Institute of Technology) in Stockholm has been nominated as the EBONE Network Operations Center. The topology of EBONE is today a pentagram with core hubs in Stockholm (Sweden), London (United Kingdom), Montpellier (France), Geneva (Switzerland) and Amsterdam (Netherlands). On-line information on EBONE is available at nic.nordu.net or archive.ripe.net in directory ebone. NSI - The Korean link has now been moved from Hawaii to NASA Ames as this solution was cheaper! The ``virtual'' upgrade of the link between Hawaii is done with compression. ANS/Merit T1/T3 - T1 traffic is now falling. 75 percent of T1 AS's has been moved to the T3 net. The T1 network continue to carry CLNP traffic. The amount of configured networks are today around 5700 of which around 4500 are announced. The RFC960 FDDI cards will be upgraded in August. The target is to have a complete cutover at the end of August. CLNP is planned to be installed during the fall. Backup of the T3 net is being planned as T1 connections between diverse ENSS`s and CNSS`s. Brazil Networking - Networking in Brazil is fostered by CNPK an analogous organization to the US NSF. The network form a core backbone with hubs are in Rio and Sao Paulo with bandwidths between 9.6 to 64 Kbps. The network is connected to the US via a 64 Kbps line from Rio to CERFnet and a 64 Kbps link between Sao Paulo and ESnet. Operational Statistics Working Group (OPSTAT) A review was made of the Internet Draft (I-D) submitted after the previous IETF. Some comments on the I-D had been received and the meeting discussed these. The ``time-section'' was renamed to ``label-section'' and extended with stop-time and a data-file-name. Clarifications will be added on how sections could be stored, one data-section per file or multiple sections per file. It was agreed that, how files are physically arranged was outside the scope of the draft. Clarification is also need for how tags should be used. A comment field will be added. Finally some editorial remarks will be included. After decided changes are included, the draft will be circulated for a final round of comments. When ready the paper will be submitted as an Informational RFC. Eric Hood, Executive Director of FARNET announced the possibility of FARNET funding the development of reference tools according to the OPSTAT model. The meeting agreed that this was important. Eric Hood 4 will make a survey of already ongoing efforts and forward the result to the OPSTAT Working Group before a decision for funding is made. User Connectivity Problems Working Group (UCP) Dan Long reported that Tom Sandoski of ConcertNet has released both a curses and X-windows interface to his freeware Trouble Ticket System on ftp.concert.net in the pub/tickets directory. There are now 50 NOC's listed in the Network Service Center Phonebook, maintained at NEARnet. Finger access is available for peer NOC's and a publicly-searchable subset of the data will be announced soon. Most of the meeting was devoted to refining the description of information that Paul Zawada has compiled for handing off tickets from one NSC to another. Several organizations have expressed interest in standardizing these handoffs. Paul will revise his ASN.1 version and Kaj Tesink will draft a specification of how to encode this information in email messages. 5