[IMR] IMR87-10.TXT Westine [Page 1] ~ OCTOBER 1987 INTERNET MONTHLY REPORTS ------------------------ The purpose of these reports is to communicate to the Internet Research Group the accomplishments, milestones reached, or problems discovered by the participating organizations. This report is for research use only, and is not for public distribution. Each organization is expected to submit a 1/2 page report on the first business day of the month describing the previous month's activities. These reports should be submitted via network mail to Ann Westine (Westine@ISI.EDU) or Karen Roubicek (Roubicek@SH.CS.NET). BBN LABORATORIES AND BBN COMMUNICATIONS CORPORATION --------------------------------------------------- WIDEBAND NETWORK The installation of a new Wideband Butterfly Gateway at Stanford University was completed at the beginning of the month. The gateway connects the Stanford campus ethernet to the Wideband Network via a terrestrial T1 circuit terminating at the gateway and at the SRI BSAT. Westine [Page 1] Internet Monthly Report OCTOBER 1987 A coast-to-coast meeting of the IAB was supported by the Wideband Network and the multimedia conferencing facilities at BBN and ISI on October 16. BSAT Release 6.0 was distributed to the Wideband Network sites this month. This new software incorporates both a user-code host interface device reset function and a round-robin control slot assignment scheme into the BSAT. These new features have been described in previous Internet monthly reports. Additionally, the new BSAT software has been adapted to run under the control of Chrysalis Operating System Release 3.0. SATNET The SATNET has performed very well through the month of October. We have had no SIMP outages and no major problems with the PSP equipment. This is the second consecutive month showing an availability of over 99% for the SIMPs from tests run by ISI. We continue to see gateway resets at the SATNET sites and are waiting for the next software release. A bug has been found in the SIMP which may be limiting the throughput of the SATNET when two channels are used. The software patch is being tested. When the patch is deployed we will be remeasuring the maximum throughput across the SATNET. Work also continues on upgrading the lines among Goonhilly, RSRE and UCL to 64 Kbps Kilostream service. The upgrade is expected to take place in late November. ARPANET USER DATABASE CUTOVER The transition of the User Database Host from the TOPS-20 machine at ISI (ISI-C) to a MicroVAX-II at BBN occurred on October 26. The transition went smoothly with no problems reported. INTERNET RESEARCH & DEVELOPMENT The major events of the month was enabling EGP fragmentation in the Butterfly and LSI-11 Gateways and updating the DDN Mailbridges to support 400 networks. We now see about active 314 networks in the Internet. The EGP fragmentation transition went well with only a few gateways reporting problems with their reassembly. There is currently serious congestion on the Arpanet to and from the EGP server gateways (ISI, BBN2, and PURDUE). We think it is caused in most part by the time it is taking these gateways (LSI- 11/23's) to generate and process EGP and GGP updates. The large Westine [Page 2] Internet Monthly Report OCTOBER 1987 processing time is due to the the growth in the number of networks that make up the Internet. We are planning to replace the LSI- 11/23 processors with LSI 11/73 processors which are significantly faster. Bob Hinden (Hinden@BBN.COM) ISI --- Internet Concepts Project Jon Postel attended the 10th Data Communication Symposium at Napa, CA, 3-7 October 1987. Jon Postel attended the Unix Workshop at Berkeley, CA, 19-20 October 1987. Paul Mockapetris attended DSAB Task Force Naming meeting at SRI, in San Jose, CA, 8 Oct 1987. The meeting concentrated on a white/yellow pages service for the future, along with general attempts to define a newer, better, and more comprehensive name service than the domain system. The general direction was toward a system which would combine multiple naming systems, but this wasn't a universally accepted view. Paul wrote an article, "Introducing Domains", Connexions the Interoperability Report, Vol., 1., No.6, October 1987. Five RFCs were published this month by Jon Postel. RFC 1021: Partridge, C., and G. Trewitt, "THE HIGH-LEVEL ENTITY MANAGEMENT SYSTEM (HEMS)", October 1987. RFC 1022: Partridge, C., and G. Trewitt, "THE HIGH-LEVEL ENTITY MANAGEMENT PROTOCOL (HEMP)", October 1987. RFC 1023: Partridge, C., and G. Trewitt, "HEMS MONITORING AND CONTROL LANGUAGEL", October 1987. RFC 1024: Partridge, C., and G. Trewitt, "HEMS VARIABLE DEFINITIONS", October 1987. RFC 1027: Carl-Mitchell, Smoot, and J. Quarterman, "USING ARP TO IMPLEMENT TRANSPARENT SUBNET GATEWAYS", October 1987. Ann Westine (Westine@ISI.EDU) Westine [Page 3] Internet Monthly Report OCTOBER 1987 Multimedia Conferencing Project The fourth packet video/voice/MMCONF teleconference meeting of the Internet Activities Board was held at ISI and BBN this month. MMCONF/Diamond was used much more successfully in this meeting, both for some prepared diagrams and for some slides brought to the meeting on paper and scanned into Diamond during the meeting (this was the first time this capability has been exercised). The problem that had caused repeated outages in the BBN gateway to the Wideband Net during a previous tele-meeting has been resolved. That repair, combined with the last-minute discovery of a long-standing MMCONF bug, improved reliability substantially. Some new features had also been added to MMCONF based on valuable comments from previous teleconferences. Video resolution was increased 25% in the horizontal dimension so the pixels now have a 1:1 aspect ratio. The packet video data rate was also increased, taking advantage of improved performance of the Wideband Net. Several short demonstrations of the conference system, including some for visiting attendees of EDUCOM '87 at USC, were also held this month. Steve Casner attended the User Interface Task Force meeting at BBN, 19-28 October 1987. Steve Casner (Casner@ISI.EDU) Brian is working on a document describing the current implementation of a document scanner system on the IBM-PC AT. Brian Hung (Hung@ISI.EDU) NSFNET Project Annette DeSchon continued working on the statistics collection programs for the NSFNET. Bob Braden attended the End2End Task Force Meeting at MIT, 21-23 October 1987. Bob Braden (Braden@ISI.EDU) Supercomputer and Workstation Communication Project Alan Katz has been spending time writing some useful tools in GNU Emacs lisp (which also runs well under X-windows). Alan wrote another ISI-Phone server user interface, developed further an EQN/Troff previewer which runs under Emacs, and have completed a references and equations numbering program which allows arbitrary naming of footnotes, references, Westine [Page 4] Internet Monthly Report OCTOBER 1987 equations, and figures and numbers them consectutively within a GNU Emacs buffer. Alan tried to make the system general enough so that it should work with Troff, Tex, Scribe, or any other directives oriented text processing language. Alan Katz (Katz@ISI.EDU) MIT-LCS ------- The network simulator project we announced last June has been going on for a few months. Up to now we have implemented the following: - At the user interface, one can use the mouse to create an arbitrary network topology, edit on it, and save it in a file for later restoration. During a simulation run, one can also use the mouse to pop up windows to observe dynamic changes of various protocol parameters in graphic display. - Two kinds of communication channel models have been coded, one emulates Ethernets, and the other, point-to-point links. - The network switch is a model of IP gateway, which forwards datagrams and sends ICMP Source Quench messages upon some given conditions. - Data sources are coded in two modules, host and transport connections, so that one is able to observe the performance of a number of transport connections that are multiplexed on the same host. At the transport level, a pseudo TCP model is being coded. - The current code measures two performance parameters, the queue length at any node (host or switch), and the transport layer round trip delay. Both of which can be observed graphically during a simulation run, and printed out on paper. A third performance parameter, the effective throughput (of a TCP connection), is being added. When the TCP module is finished, we will be able to make some interesting demonstrations. A long waiting list of simulation tests has been created, which includes observing the Westine [Page 5] Internet Monthly Report OCTOBER 1987 oscillating behavior of network traffic under window flow control, doing some quantitative study on the suggested TCP slow-start mechanism, etc. More functional modules will also soon be added to the simulator for further studies, such as comparing window flow control with rate flow control. Lixia Zhang (Lixia@XX.LCS.MIT.EDU) MITRE Corporation The objective of the MITRE Internet Engineering program is twofold: 1) to address internet level performance issues for the DoD internet and 2) to address the interoperability between the DoD and OSI protocol suites as support for a planned transition from DoD to OSI protocols. During October, we continued work in the following areas: 1. Documentation Availability For those folks that asked for MITRE documentation, as soon as MITRE and OSD public release are achieved, I will send the documentation to you. Sorry, but the process could take a few months. If anyone wants to set up conversations with the respective authors, let me know and I will direct you to them. 2. VTP Implementation We continued work on the Basic Class VT implementation. We have upgraded the VT interface to ISODE Phase 3.0 and are in the process of changing software to reflect the current state machine defined in the second draft DIS. This year's work will also include implementation of forms mode and synchronous mode capabilities. We are providing a comparison of Honeywell VIP functions and VT forms mode functions for the November ANSI X3T5.5 meeting. 3. Performance Analysis We added spectral analysis techniques to data obtained via internet performance measurements and congestion control simulation. The first application is analyzing round trip times for periodicities in their variation. 4. Landmark Routing We installed a network simulation package, OPNET developed by MIL 3, INC., which is ported to an BSD 4.2 Apollo. We are using the simulator to develop the state machine for the landmark routing Westine [Page 6] Internet Monthly Report OCTOBER 1987 protocol specification. 5. Congestion Control We applied nonparametric statistical analysis to the set of congestion control simulation results. We also used the boolean minimization program "expressso" to categorize performance ranges of simulation results obtained by varying the parameters to the experiments. These results will be presented at the November IETF. 6. DoD/ISO Transition Plan We continued redrafting the Plan to incorporate comments from the initial government review. The Plan is scheduled for 60 day formal DoD review at the end of November. 7. FTP/FTAM Bridge We upgraded the implementation to use the new Phase 3.0 ISODE release. Plans are to offer the bridge to the network community for experimental use. Ann Whitaker (Whitaker@Gateway.Mitre.Org) NTA & NDRE ---------- 1. The Nordic telecommunications administrations have decided to turn the antenna dish at Tanum, in use for SATNET, from the current position of 325 degrees to 342 degrees. The turn will be performed in January/February. The decision was made without any consultation or prior warning. The central administration has admitted their mistake and try to come up with a workable solution, which seems to be a 9.6 kb/s land line from the Butterfly gateway at NTA-RE to the Goonhilly SIMP (plugging into the second host port). 2. I have collected and sent over to Jon Reinhold of BBN information on Norwegian-produced satellite groundstation equipment which may be used in a possible upgraded version of SATNET. We will work more on this matter, so that I can bring a more definit plan to the January meeting in San Diego. 3. A Norwegian research network, covering all major research and educational sites, is in its final planning stage. It will be sponsored by Dept. of Education, Dept. of Industry and NTA. Westine [Page 7] Internet Monthly Report OCTOBER 1987 The network will initially support two sets of standards, X.25 and TCP/IP, but will gradually be converted to full ISO standards as they are made available. It will utilize Local Area Networks, private X.25 switches, leased lines and the public X.25 network. The implementation will start in the middle of next year. 4. SATNET and Butterfly seems to behave very well, but connectivity into ARPA-land is not much to be proud of. But the combination of the old PDP-11 gateway and the Butterfly, supposed to establish connectivity between our Pronet Ring and the rest of the world including Norway, is giving us much grief. Most of the time the routing table in the PDP is empty, thus no connectivity. We have many users and machines on the Ring, including NDRE and NORSAR, thus connectivity is of great importance. BBN has been examining the problem without finding the source of the problem. But they have promised to give priority to writing a Pronet driver for the Butterfly, so that we can do away with the PDP-11. Paal Spilling (Paal@TOR.NTA.NO) SRI --- Internet Research In October, Jose Garcia presented four papers at MILCOM'87, Zaw- Sing Su attended IETF Open Routing Working Group meeting at BBN and contributed to the writing of a draft report, "Requirements for Inter-Autonomous Systems Routing". Zaw-Sing Su UCL --- Research: Work continues on ISO protocols using ISODE. A specification for Layer Management of Transport, session and ACSE has been written, and implementation has been started. Work on a Diamond to pure ODA document conversion program continues. Westine [Page 8] Internet Monthly Report OCTOBER 1987 A ROS/X.400 based pilot directory service including a picture server has gone into trial. Infrastructure: UCL has recently been seeing very poor connectivity to remote networks. We can see the US end of SATNET, but routes to networks beyond that appear to fluctuate in and out of reachability very rapidly. Relaying file transfers via hosts on the East Coast makes access feasible. The question must be asked that if a human can find a stable relay route (up for over 1 hour), why can the gateways not find a stable IP route? John Crowcroft (jon@CS.UCL.AC.UK) UNIVERSITY OF DELAWARE ---------------------- 1. The massive disruption of the NSFNET community reported last month has subsided somewhat, but continues at disturbing levels. While traffic for various nets was redistributed over the three gateways at Pittsburgh Supercomputing Center and U Maryland, virtual-circuit thrashing is now seen at all three gateways, resulting in occasional broken connections between the gateways and other ARPANET hosts and gateways. The problem has to do with a Unix driver for the ACC 5250 X.25 interface and its managment of resources when the maximum number of virtual circuits (64) is reached. 2. Investigation and experiment on Internet congestion control continues. An analysis of the NSFNET Backbone logs showed the new congestion-control (more properly congestion-avoidance) mechanisms are working as expected. However, also as expected, few hosts pay any attention to the ICMP Source Quench messages generated, in some cases at levels of a few percent of the aggregate traffic. Since these messages are logged, it is in principle easy to identify persistant abusers which ignore these things; however, this remains a task for the Backbone managers to pursue. 3. Several new nets were added to the Backbone routing tables during the month, bringing the total to about 95, which is about one-third of the total nets known to the core system. Due to causes yet to be found, some of these nets become quite unstable, with the effect that delays frequently oscillate wildly and loops occasionally form. Thorough review and Westine [Page 9] Internet Monthly Report OCTOBER 1987 testing revealed the spike-suppression and loop-avoidance mechanisms of the routing algorithm could be improved in several minor ways. The neccessary changes were implemented and tested, then distributed to the Backbone and related communities. 4. New features were added to the Backbone routing algorithm to allow a "floating default" where traffic for ARPANET and nets not reachable in the NSFNET community can flow out the nearest gateway, rather than a static route via a gateway clear across the country. 5. Development continues on the Dissimilar Gateway Protocol (DGP). Most of the intricate design issues have been resolved. Mike Little is working on a preliminary specification document due in mid-November. 6. Mike Minnich attended the INENG meeting in Boulder. Dave Mills attended the IAB meeting and also the Tenth DATACOM meeting at Napa Valley, where he presented briefings on the NSFNET Backbone, congestioncontrol mechanisms and routing algorithms. 7. Planning continues for the INARC Workshop, which has by popular demand (an understatement) been rescheduled for 17-18 December at BBN. One of the nice things is lively interest beyond the usual Internet buzzards, in particular the UK and US industrial research communities. See the INARC report elsewhere for further information. Dave Mills (Mills@UDEL.EDU) Westine [Page 10] Internet Monthly Report OCTOBER 1987 NSF NETWORKING -------------- NSF NETWORKING UCAR/BBN LABS NSF NETWORK SERVICE CENTER (NNSC) The RFCs that define the High-Level Entity Management System (HEMS) are now available from the Network Information Center in the directory at SRI-NIC.ARPA. The second issue of the NSF Network News has been completed and will be distributed in mid-November. Craig Partridge attended the End-to-End Task Force Meeting and represented the NNSC with Karen Roubicek at the meeting of the Internics Working Group of the Internet Engineering Task Force. ***PLEASE NOTE***The new NNSC hotline number has been changed to (617) 873-3400. By Karen Roubicek (roubicek@nnsc.nsf.net) NSFNET BACKBONE SITES CORNELL UNIVERSITY THEORY CENTER New fuzzball software was installed which is expected to dramatically improve NSFNet behavior (see Dave Mills's report). Unfortunately there was no way this software could be tested in an environment anywhere near as complex and stressful as the NSFNet backbone, so we are still uncovering problems and improving it. Also the new software does not yet work in perfect harmony with the gatedaemons that form the protective wall around the backbone, although we have made great steps in that direction (new fuzzball and gatedaemon fixes are currently under test and about to be deployed; even more fixes will follow). The result of all of this is that there were some routing problems which reduced NSFNet robustness (although not nearly as much as we had feared). In addition our Arpanet connection needs grew far beyond the limit of 64 Virtual Circuits in our main DDN X.25 interface to Arpanet, so that NSFNet network members would appear reachable to some Arpanet sites but not others, at various times. We have redistributed which NSFNet->ARPANet gateways announce which NSFNet component networks with what EGP metrics, and that problem seems to be under control for the moment. We have also decreased the virtual circuit idle timeout on our busiest gateway from 10 minutes to 3 minutes on an experimental basis. The NSFNET community currently consists of approximately 120 Westine [Page 11] Internet Monthly Report OCTOBER 1987 networks (the number varies with policy and operational changes). Mark Fedor has gone to work exclusively on NYSERNET issues. NSFNET Gatedaemon configurations will be handled by Cornell's Network Information and Support Center (send questions to oros@tcgould.tn.cornell.edu). Mark will take most of his consulting with him. For gatedaemon development we have a good person starting November 30th; until then we will be covering the most important fixes as best we can. Summary statistics from Doug Elias (more are available, of course): NSFNET TRAFFIC REPORT Period: Sept., 1987 Total Traffic Figures Between Sites Ethernet Input 102868099 71756593 Output 102197912 70629065 In+Out 205066011 142385658 Grand 347451669 By Scott Brim (swb@tcgould.tn.cornell.edu) UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN Midnet's networks have been announced to the fuzzball hence to the Internet. Midnet's routing is still not up to snuff due to inability of P4200 7.3 software to deal with multiple subnetted networks. Version 7.4 should fix the problem and be available by 15 Nov. Rerouting of the NCAR-UIUC link directly to Urbana (bypassing an outrageously noisy T1-broadband cable hop from Chicago) was done. We will be placing an order immediately, with delivery due in 45 working days, to reterminate the SDSC-UIUC link to SDSC-PSC. This will cut the diameter of the NSFnet backbone to 2 from 3. Our test window for ARPAnet has come and gone. DDN has promised to have two lines to us by 5 Dec to correspond with the installation of an additional PSN at Madison. By Ed Krol (krol@uxc.cso.uiuc.edu) JOHN VON NEUMANN NATIONAL SUPERCOMPUTER CENTER This report is a summary of the JVNCNET Monthly Report. A copy of the full report can be obtained by sending a request to "JVNCNET- Westine [Page 12] Internet Monthly Report OCTOBER 1987 nic@jvnca.csc.org". The purpose of the report is to inform the JVNC Consortium and the JVNCNET network members as well as the Internet Community of the monthly activities and status of the JVNCNET network. The data used on this report is collected using a number of techniques developed at JVNC, together with data from the JVNC operations group. The NSFNET data is obtained from the NSFNET NOC (Network Operations Center). Network Statistics: ------------------ 1.- links and gateways availability better than 95.43 % (last month 90.46%) 2.- high speed lines status all available 100% 3.- satellite links status no significant events 4.- traffic subnet 50 (better than): 145,685,725 packets in one direction JVNC Supercomputer traffic/total JVNCNET =~ 46.26% 5.- Routing maximum degree of unstability 600 kernel route changes per 10 minutes in bursts. Almost all routing changes coming from the NSFNET backbone. Projects: -------- 1.- Network Monitoring: Status: On going Working on providing backup monitoring capabilities to netsc so we can keep the network monitored even when our primary monitoring Westine [Page 13] Internet Monthly Report OCTOBER 1987 server is down. 2.- Network Characterization: Status: On going The Network Characterization program is directed towards determining the parameters that characterize the JVNCNET network's diverse type of services. This effort will be utilized not only as a research subject but to find/predict network bottlenecks and problems before they are obvious to the end users. This task started two weeks ago with the collection of data and the automation of the collection, and will continue with the study of the characteristics and patterns that distinguish each point to point lines whether T1, 56kpbs or satellite. The results will be available to the community. 3.- Traffic Analysis: Status: On going The traffic data is currently being collected from JVNCA, JVNCB, and the fuzzball automatically. This will be extended to the other VAXs on the same ethernet and the other routers on the network. At the same time the data will be compiled per campus. 4.- Routing Analysis Package: Status: On going This package, now under development, will permit the operators keep track of the routing stability of the network, as well as the current status of the routing. Part of it is now providing the network operations staff with routing stability indicators. Comments: -------- The gateways were available more time than the previous month. The amount of traffic passed through the major ethernet in one direction was higher than last month (even though there are missing observations of the data, thus representing a low bound). The characteristics of the traffic for this month (not counting the effect that not all the data was available) shows that the "Supercomputer traffic" for the month was higher than the previous month. The NSFNET backbone suffered from serious routing problems making our internal routing as unstable as the one on the NSFNET backbone (due to the fact that we all share the same routing flat space). The routing behavior is being watched very closely to Westine [Page 14] Internet Monthly Report OCTOBER 1987 determine some known anomalies and some new ones, as well as the interaction between the NSFNET backbone and the JVNCNET network. The high speed lines have all behaved well this month. The satellite system had no major problems except for a glitch with the modulator at JVNC which was quickly detected (by JVNC staff) and reset. PSN still not connected, BBN sent a technician a couple of times, but he was unable to bring up the PSN for a number of reasons. Information: ----------- JVNCNET NOC: "net@jvnca.csc.org" (JVNCNET Network Operations Center) JVNCNET NIC: "JVNCNET-nic@jvnca.csc.org" (JVNCNET information) JVNCNET manager: "heker@jvnca.csc.org" * CSC Consortium: Princeton University, MIT, Harvard, Brown, University of Pennsylvania, Rutgers, IAS, Columbia, University of Rochester, NYU, Penn State, University of Arizona, University of Colorado. ** NRAC (Newark Remote Access): New Jersey Institute of Technology (NJIT), Stevens Institute of Technology, University of Medicine and Dentistry of New Jersey (UMDNJ). By Sergio Heker (heker@jvnca.csc.org) NATIONAL CENTER FOR ATMOSPHERIC RESEARCH & UNIVERSITY SATELLITE NETWORK PROJECT The USAN connection to NRL (Naval Research Lab) in Washington has been installed, finally, and is to undergo acceptance testing the first half of November. A network connection to the NASA Science Center at NASA Ames has been put into place. The connection is via Proteon p4200 routers and a 56Kbit land line. The Proteon router drops onto the USAN (128.116) backbone at NCAR. By Don Morris (morris@scdsw1.ucar.edu) Westine [Page 15] Internet Monthly Report OCTOBER 1987 PITTSBURGH SUPERCOMPUTING CENTER No report received. SAN DIEGO SUPERCOMPUTER CENTER We have developed software that allows the encapsulation of IP traffic within our consortium network, SDSCNET, which uses the MFENET protocols. It relies on the remote VMS VAX running either SRI or TWG for its local IP support. We hope to add support for other packages in the future. Additional, we are using much the same new software to move IP traffic over DECNET links, again assuming that the VAX's are using SRI or TWG. Our PSN's first trunk line (to USC/ISI) was installed on the 29th. With luck, the line should be tested in another week. Perhaps we will have the PSN moving packets by year end. The Proteon with 7.3b continues to be reliable. We are looking forward to 7.4 and another 512k for the box. We have had no change in our line configuration this past month. By Paul Love (loveep@sds.sdsc.edu) Westine [Page 16] Internet Monthly Report OCTOBER 1987 29 Oct 1987 San Diego Supercomputer Center Internet Connections SDSC Ethernet ____________ || | NSFnet |----to NCAR; 56kb ||--------| Fuzzball |----to NASA/AMES; 56kb (coming '88) S || |__________|----to UIUC; 56kb E || ____________ V || | Bridge |----to Industrial Participant; 56kb E ||--------| GS/3 | R || |__________|----to UC Irvine; 56kb (coming ~1 Feb) A || ____________ L-------|| | ACC 5250 | : ||--------|in micrVAX|---->to ARPANET PSN #26 S : || |__________| at SDSC D : || (coming 1 Jan) S : || C : || : || H-------|| ____________ O || | Proteon |----to UC Santa Barbara; 56kb S ||--------| p4200 |----to UC Berkeley; 56kb T || |__________|----to Industrial Part.; 56kb S || | | (coming ~1988) || | |------to Salk Inst; T1** || | **DECNET also carried & || |----|| || ____________ || ____________ || | DEC LAN | || | Unix Sys | to MILNET 9 ||---| Bridge100|---|| ______________ ||---| SDCSVAX |-->PSN @ NOSC 5-------|| |__________| || | UB Buffered| || |__________| at 56kb : || (filters local ||---| Reperater |---|| ____________ M : || traffic plus all|| |____________| || | Unix Sys | to UCDLA A : || broadcast pkts) || ||---| SDCC18 |-->PSN (Net C-------|| || UCSD |__________| # 31) S || (DELNI) Ethernet || (UB broadband) || ____________ || | Apollo | ||--------| 580 |--->SDSC Apollo "ring" of Apollo 3000's, etc || |__________| || Westine [Page 17] Internet Monthly Report OCTOBER 1987 Internet Network Numbers School/Institution Network Number Domain San Diego Supercomputer 192.12.207 sdsc.edu SDSC Apollo Ring 192.31.21 sdsc.edu UC Santa Barbara 128.111 ucsb.edu UC San Diego 128.54 ucsd.edu UC Irvine* (temp. path) 128.200 uci.edu UC Riverside* (pending) 192.031.146 ucr.edu Salk Institute 192.031.153 sdsc.edu Ind. Part # 1 (GS/3) not registered Ind. Part # 2 (p4200) 192.031.248 .arpa * Over DECNET Bridge NSFNET REGIONAL AFFILIATED & CONSORTIUM NETWORKS BARRNET (No report received) JVNCNET (Refer to JVNNSC backbone report) MERIT At the Merit Network central site in Ann Arbor, recent TCP/IP development work has focused on the Telnet-to-Hermes gateway function, with work continuing on congestion control in Merit's own nodes. University of Michigan staff in Ann Arbor have completed installing UDP into MTS, a major Merit host system; have the implementation well under way for the HIM (Host Interface Module) that will pass TCP/IP into MTS -- protocols supported initially will include FTP and Telnet; and have been working on porting BIND into the MTS environment. On the U-M campus, Merit's Ethernet IP support is now in production use for bridging Appletalk networks using the Kinetics FastPath hardware. Merit member Michigan State University is near completion of the installation of their own direct land line to the ARPA IMP at Purdue. They are also ready to test switching their direct Fuzzball link into Merit/Ann Arbor over to the use of Merit's encapsulated IP over the normal Merit links. In Kalamazoo, Merit member Western Michigan University has placed an order for a MicroVax/Ultrix system to use as a DECNET to Internet gateway; in addition, WMU is beginning to use the Merit Ethernet support for IP on its campus. By Christine Wendt (christine_wendt@um.cc.umich.edu) Westine [Page 18] Internet Monthly Report OCTOBER 1987 MIDNET In the last week MIDNet has become completely operational. All of the member sites are speaking to one another and speaking to the ethernet at Illinois at Champaign-Urbana. We have not yet been announced to the Internet at large but we expect this to happen fairly soon. There are already researchers at several campuses starting to do work at NCSA and others waiting to get to other sites where they have time. We at UNL are planning to take an informational road show out to the members of MIDNet starting in November. We plan to put on meetings at each of the member campuses where we will discuss with local users what they can use MIDNet for and how they can use it. We hope these meetings will prove helpful to everyone. By Dale Finkelson (dmf@relay.cs.net) NORTHWESTNET 1. A copy of the network topology is shown below. The first links should be operational in November; we hope that the whole network will be running early next year. We are using Proteon routers at all the nodes. Westine [Page 19] Internet Monthly Report OCTOBER 1987 --------------------------------------------------------------------- Fairbanks (Alaska State Sys.) | | 56 (satellite) | | Seattle 56 Bellevue 56 Pullman 56 Moscow (Univ. of -------- (Boeing ---- ----- (Washington ------ (Univ. of Washington) Comp. Serv.) State U.) Idaho) | | | 56 19.2 | | | | | Portland Bozeman (Oregon Grad. (Montana Center) State U.) | | | 56 19.2 | | | | | Corvallis Fargo (Oregon ................... (N. Dakota State U.) : State Sys) | : | | 56 : | | : | | : | Eugene : 19.2 | (Univ. of -----------------------------------------------------+ Oregon) : : existing experimental : 56/224 satellite link : : San Diego :......... Boulder (SDSC/ (NCAR/ NSFNET) NSFNET) All speeds, e.g. 56, are in kbits/sec. The existing MFE satellite link between Seattle and San Diego is not shown. ---------------------------------------------------------------------- Westine [Page 20] Internet Monthly Report OCTOBER 1987 2. The link between Northwestnet and NSFNET has not yet been determined. We have requested NSF to provide funding for a terrestial 56 kb/s link between Seattle (University of Washington) and San Diego (SDSC). This link would also replace the existing SDSC satellite link between Seattle and San Diego, which uses the MFE protocols. We also recommended to retain the experimental VITAlink connection between Corvallis and Boulder. 3. By the time you read this, the Management and Technical Committees of Northwestnet will have held a two-day conference at the San Diego Supercomputer Center, October 26 and 27. The first day is devoted to an overview of supercomputing and networking issues for faculty and administrators from the participating institutions. 4. We have started preliminary discussion with institutions of higher education in South Dakota regarding the possibility for them to join Northwestnet. By Hellmut Golde (GOLDE@CS.WASHINGTON.EDU or GOLDE@UWACDC.BITNET) NYSERNET As of 1 November 1987, NYSERNet had the following topology: Clarkson Syracuse--+ | | | Rochester--------Cornell---------RPI---Albany | | | Buffalo...Fredonia | | | ....Oswego | | | | | Binghamton | +-------- | ------StonyBrook | | | | | | | | CUNY--NYTEL/NSMAC--Columbia======NYU-+ | |\ | | | / | | | | \ NYTEL/GC | NYNEX/S&T / | | | | \ BNL / | | | | \ / | | | | +-------------Rockefeller | | | | | | | +------------------------------+ | | | +-------------POLY-------------------+ ==== T1 \ | / 56kbits .... 9.6kbits Westine [Page 21] Internet Monthly Report OCTOBER 1987 A new site, SUNY Oswego, was connected to NYSERNet at 9.6kbits this month. The NYSERNet Network Information and Support Center opened outside of Albany, NY combining RPI/NIC and Cornell/NOC services with permanent full-time staff By Marty Schoffstall (schoff@nisc.nyser.net) SDSCNET (Refer to SDSC backbone report) SESQUINET The complete initially proposed SesquiNet configuration has been operational now for three months. The following campus networks are being served, and are advertised via EGP to the core: Baylor College of Medicine 128.249 Houston Area Research Center 192.31.87 Rice University 128.42 Texas A&M University 128.194 Texas Southern University 192.31.101 and the University of Houston 129.7 Although the network is not very heavily loaded, we have still experienced no gateway failures. We have now done some analysis of the reliability of the network, based on logs written by our monitoring program. For each of the sites listed below, we note the number of (unscheduled) failures and the percent of availability (percentage of time the site was up, deducting scheduled down time). -------------------------------- Site Failures Available -------------------------------- BCM 0 100.0 % HARC 5 93.7 % RU 0 100.0 % TAMU 2 99.2 % TSU 2 98.9 % UH 2 98.9 % -------------------------------- While some of the problems were due to leased lines, most of the unscheduled downtime was due to machine room power problems. Westine [Page 22] Internet Monthly Report OCTOBER 1987 We continue to await the completion of the connection from NSFNET/NCAR to SesquiNet/Rice via fuzzballs at the two sites. By Guy Almes (almes@rice.edu) SURANET The number of nodes in SURANET continues to increase. The following nodes are presently connected and functioning: 1)University of Maryland 2)George Washington University 3)Virginia Tech 4)TUCC (NC State, UNC, Duke) 5)Clemson University 6)Georgia Tech 7)U of Alabama (Birmingham) 8)University of Tennessee 9)University of Kentucky 10)University of Delaware 11)National Science Foundation 12)Florida State University 13)George Mason University 14)Gallaudet University 15)Catholic University of America 16)University of Virginia 17)CEBAF (Continuous Electron Beam Accelerator) 18)Johns Hopkins University 19)NASA/Goddard 20)Vanderbilt University 21)Naval Research Laboratory 21)Louisiana State University 22)University of Florida All of the above lines are 56kbs. The line to NASA/Goddard will be upgraded to T-1 within a few days. A T-1 line from the National Bureau of Standards will be installed very shortly connecting the Bureau to SURANET via the University of Maryland. In addition the National Radio Astronomy Laboratory will shortly connect to SURANET via an ethernet connection to the University of Virginia and the College of William and Mary will connect to SURANET via a 56kbs line to CEBAF. Westine [Page 23] Internet Monthly Report OCTOBER 1987 Videotapes of the two SURANET meetings are still available. The 6 two hour tapes of the Feb meeting are available for $91 and the 7 two hour tapes of the July meeting are available for $105. If you would like to purchase copies of one, or both, contact HAHN@UMDC.UMD.EDU. By Jack Hahn (HAHN%UMDC.BITNET@wiscvm.wisc.edu) WESTNET 1. Equipment, including CSU/DSU's and IP Gateways, has been ordered for the implementation of Phase II of Westnet. A total of five IP Gateways will be installed midway through November at the locations of: New Mexico Institute of Mining and Technology, New Mexico State University, New Mexico Technet, University of New Mexico, and the University of Wyoming. 2. A 9.6 kbps circuit has been ordered to connect the University of Wyoming to Colorado State University, and will be installed Dec. 1. The circuit will be upgraded to 56 kbps as soon as this service becomes available into Laramie. New Mexico Technet has applied to the AT&T Foundation for cost sharing of a circuit between Albuquerque and NCAR, until the FY'88 money becomes available. Until then, New Mexico Technet is providing 56 kbps circuits within the state of New Mexico as cost sharing. 3. A workshop for Westnet Technical Personnel is being organized to be presented from Nov. 9 to 11 at the University of Colorado at Boulder. Personnel from 16 Westnet sites will be in attendance. The first two days of the workshop will consist of lectures, and describe the features of TCP/IP and the Internet. The final half day will consist of a handson session with the cisco IP Gateways. At that time, the Gateways will be configured for the specific campuses which will receive the devices. 4. Brigham Young University and the University of Utah have requested to connect to NSFNET via Westnet. We are currently exploring the costs and administrative structure of this arrangement. Westine [Page 24] Internet Monthly Report OCTOBER 1987 5. Joe Choy has granted permission for Westnet to plug circuits from Salt Lake City and Albuquerque into the NCAR cisco Gateway. Westnet will provide a 56 kbps board (2 ports)to effect these connections, and an upgrade from the 68000 CPU board to the 68020 CPU board to enhance the performance of these connections. By Pat Burns (pburns%csugreen.bitnet@WISCVM.WISC.EDU) Westine [Page 25] Internet Monthly Report OCTOBER 1987 TASK FORCE REPORTS ------------------ APPLICATIONS -- USER INTERFACE The task force held its second meeting October 20-21 at BBN Labs. The first topic of discussion were some minor revisions to our charter. First, we anticipate making contributions at three different levels: 1) reference models, 2) architectures, and 3) implementations. (For comparison, consider the OSI Reference Model, the various network architectures that conform to that model, and the various implementations of each such architecture.) Second, our "current focus" has been broadened somewhat to be "multimedia communications (including real-time, computer-supported teleconferencing) and user interface architecture". In the spirit of multimedia communications, much of the meeting was spent discussing the strawman voice server, as put forth by Chris Schmandt. There being general agreement on the overall architecture and level of functionality currently offered, several members committed to acquiring the requisite hardware and software and begin experimenting--with alternative architectures for voice storage (e.g. voice server-based disk vs. workstation disk vs. network-based file server) and alternative phone interfaces (e.g. analog vs. commercial PBXs vs. ISDN), in particular. While we hope to produce a set of RFCs shortly, interested parties should tune in (or stay tuned) to the UI-INTEREST distribution list. Other items of discussion included an update on the video component of the Wideband-based teleconferencing system, mechanisms for conversation control and voice-editing in the Etherphone project, and the human factors of voice communication. More details will appear in the minutes of the meeting, to be posted to UI-INTEREST. The next meeting is tentatively scheduled for February 9-10 at ISI. User interface architecture will be a major topic of discussion. Keith Lantz (Lantz@SCORE.STANFORD.EDU) Westine [Page 26] Internet Monthly Report OCTOBER 1987 AUTONOMOUS NETWORKS Nothing to report this month. Deborah Estrin (Estrin@OBERON.USC.EDU) END-TO-END SERVICES No progress to report this month. Bob Braden (Braden@ISI.EDU) INTERNET ARCHITECTURE This report is largely a repeat of last month's report and is included primarily because of a change from the original 17-18 November meeting dates to 17-18 December. The meeting place is unchanged. Please note the revised and clarified wording in the announcement itself. The Internet Architecture Task Force (INARC) studies technical issues in the evolution of the Internet from its present architectural model to new models appropriate for very large, very fast internets of the future. It is organized as a recurring workshop where researchers, designers and implementors can discuss novel ideas and experiences without limitation to the architecture and engineering of the present Internet. The output of this effort represents advance planning for a next-generation internet, as well as fresh insights into the problems of the current one. The INARC is planning a two-day retreat/workshop for 17-18 December at BBN to discuss a fresh start on advanced internet concepts and issues. The agenda for this meeting will be to explore architecture and engineering issues in the design of a next-generation internet system. The format will consist of invited presentations on selected topics followed by a general discussion on related issues. Written contributions of suitable format and content will be submitted for publication in the ACM Computer Communication Review. In order to have the most stimulating discussion possible, the INARC is expanding the list of invitees to include any researchers with agenda to plow, axe to grind, sword to wield or any other useful instrument for that matter. While not a precondition for admission, participants are encouraged to Westine [Page 27] Internet Monthly Report OCTOBER 1987 contribute concise presentations, either written or oral (fifteen to thirty minutes), in electronic form to mills@udel.edu or in hardcopy form to Dr. David L. Mills Electrical Engineering Department University of Delaware Newark, DE 19716 (302) 451-6534 or 737-9211 Speakers will be selected on the basis of quality, relevance and interest. Every effort will be made to accomodate all participants that wish to attend; however, participants are asked to contact the chairman by electronic mail or telephone at least a week in advance to confirm their intention to attend. Following is a list of possible areas and issues of interest to the community. Readers are invited to submit additions, deletions and amendments. 1. How should the next-generation internet be structured, as a network of internets, an internet of internets or both or neither? Do we need a hierarchy of internets? Can/must the present Internet become a component of this hierarchy? 2. What routing paradigms will be appropriate for the new internet? Will the use of thinly populated routing agents be preferred over pervasive routing data distribution? Can innovative object-oriented source routing mechanisms help in reducing the impact of huge, rapidly changing data bases? 3. Can we get a handle on the issues involved in policy- based routing? Can a set of standard route restrictions (socioecononic, technopolitic or bogonmetric) be developed at reasonable cost that fit an acceptable administrational framework (with help from the Autonomous Networks Task Force)? How can we rationalize these issues with network control and access-control issues? 4. How do we handle the expected profusion of routing data? Should it be hierarchical or flat? Should it be partitioned on the basis of use, service or administrative organization? Can it be made very dynamic, at least for some fraction of clients, to support mobile hosts? Can it be made very robust in the face of hackers, earthquakes and martians? Westine [Page 28] Internet Monthly Report OCTOBER 1987 5. Should we make a new effort to erase intrinsic route- binding in the existing addressing mechanism of the Internet IP address and ISO NSAP address? Can we evolve extrinsic binding mechanisms that are fast enough, cheap enough and large enough to be useful on an internet basis? 6. Must constraints on the size and speed of the next- generation internet be imposed? What assumptions scale on the delay, bandwidth and cost of the network components (networks and gateways) and what assumptions do not? 7. What kind of techniques will be necessary to accellerate reliable transport service from present speeds in the low megabit range to speeds in the FDDI range (low hundreds of megabits)? Can present checksum, window and backward- correction (ARQ) schemes be evolved for this service, or should we shift emphasis to forward-correction (FEC) and streaming schemes. 8. What will the internet switch architecture be like? Where will the performance bottlenecks likely be? What constraints on physical, link and network-layer protocols will be advisable in order to support the fastest speeds? Is it possible to build a range of switches running from low-cost, low-performance to high-cost, high-performance? 9. What form should a comprehensive congestion-control mechanism take? Should it be based on explicit or implicit resource binding? Should it be global in scope? Should it operate on flows, volumes or some other traffic characteristic? 10. Do we understand the technical issues involved with service-oriented routing, such as schedule-to-deadline, multiple access/multiple destination, delay/throughput reservation and resource binding? How can these issues be coupled with effective congestion-control mechanisms? 11. What will be the relative importance of delay-based versus flow-based service specifications to the client population? How will this affect the architecture and design? Can the design be made flexible enough to provide a range of services at acceptable cost? If so, can the internet operation setpoint be varied, automatically or manually, to adapt to different regimes quickly and with acceptable thrashing? Westine [Page 29] Internet Monthly Report OCTOBER 1987 12. What should the next-generation internet header look like? Should it have a variable-length format or fixed- length format? How should options, fragmentation and lifetime be structured? Should source routing or encapsulation be an intrinsic or derived feature of the architecture? 13. What advice can we give to other task forces on the impact of the next-generation internet in their areas of study? What research agenda, if any, should we propose to the various NSF, DARPA and other agencies? What advice can we give these agencies on the importance, level of effort and probablity of success of the agenda to their current missions? Dave Mills (Mills@HUEY.UDEL.EDU) INTERNET ENGINEERING 1) Several new working groups have been organized or formalized since the July meeting at MITRE. They are: - Open Systems Routing (Hinden, BBN) - Open Systems Internet Operations Center (Case, UTK) - Performance and Congestion Control (Stine, Mitre) - Open IGP (Petry, UMD) - InterNICs (Feinler, SRI-NIC) - Internet Host Requirements (Gross, Mitre) - Internet Authentication Protocol (Schoffstall, RPI) 2) The most recent IETF meeting was (Mon-Weds, Nov 2-4) at NCAR in Boulder, Colorado. The meeting was hosted by Don Morris. He did an excellent job of handling all local arrangements. He arranged for local Internet connectivity for reading mail and doing local document editing. 3) The agenda for the 3 day (Nov 2-4) IETF meeting is below. The basic format is: 1) Working Group meetings for the first 1.5 days, 2) Presentations and network reports on Tuesday afternoon and 3) Presentations, Working Group reports and discussion on Wednesday. The several new working groups that will be introduced and discussed during the opening plenary on Monday. All groups Westine [Page 30] Internet Monthly Report OCTOBER 1987 will be open. If the chair feels that more detailed discussion is required to achieve results, she/he has the option of calling meetings between regular IETF meetings. Agenda for the November 2-4 IETF meeting at NCAR MONDAY, NOVEMBER 2 - Opening Plenary (local arrangements, Discussion of IETF format, overview of new working groups) - Working Group meetings convene - Open Systems Routing (Hinden, BBN) - Short Term Routing, Old Business (Hedrick, Rutgers) - Open Systems Internet Operations Center (Case, UTK) - Performance and Congestion Control (Stine, Mitre) - Open IGP (Petry, UMD) - Domain Issues (Lottor/Stahl, SRI-NIC) (Lunch and Breaks scheduled by Chairs) - Recess at 5:00pm TUESDAY, NOVEMBER 3 Morning - Opening Plenary - Working Group meetings convene - Internet Host Requirements (Gross, Mitre) - EGP3 (Gardner, BBN) - Internet Authentication Protocol (Schoffstall, RPI) - InterNICs (Feinler, SRI-NIC) - Short-Term Routing, New Business (Hedrick, Rutgers) Afternoon - Management/Monitoring Working Group Report (Partridge, BBN) - SGMP Status and Demonstration (Case, UTK) - NSFNET Report (Wolff, NSF) - BBN Report (Hinden/Gardner, BBN) - Recess at 5:00pm WEDNESDAY, NOVEMBER 4 Morning - Opening Plenary Westine [Page 31] Internet Monthly Report OCTOBER 1987 - IP over 802.X (Perkins, CMU) - Congestion Control Simulation Results (Stine, Mitre) - Recent Congestion Control Efforts for 4.2/4.3BSD (Van Jacobson, LBL) Afternoon - Working Group Reports and Discussion - Open Systems Routing (Hinden, BBN) - Short Term Routing (Hedrick, Rutgers) - InterNICs (Feinler, SRI-NIC) - Open Systems Internet Operations Center (Case, UTK) - Performance and Congestion Control (Stine, Mitre) - Open IGP (Petry, UMD) - Internet Host Requirements (Gross, Mitre) - Domains (Lottor/Stahl, SRI-NIC) - EGP3 (Gardner, BBN) - Internet Authentication Protocol (Schoffstall, RPI) - Concluding Discussion, announce next meeting. - Adjourn 4) Proceedings for the July IETF meeting are available and a limited number of copies will be distributed at the November meeting. Allison Mankin (Mitre) worte the meeting notes and did the bulk of the work in compiling the Proceedings. She did an excellent job on a sizable document (the July Proceedings are approximately 280 pages!). She had assistance from several folks (notably Coleman Blake, Mitre) who contributed writeups of various presentations. The Proceedings include the reports from the working group meeting in July. Due to the size of the Proceedings, only a limited number of copies will be carried to the November meeting for distribution. An arrangement is being worked out in which paper copies can be obtained by U.S. mail. 5) A number of the working groups have produced documents. Some of these documents are planned to be submitted as propsed RFCs after careful review. As a bookkeeping measure, we have established an IETF document numbering series and online copies of the documents will be kept in the IETF directory at the NIC. An announcement will be made when this has been completed. 6) Anyone wishing to receive more information about IETF, such as obtaining copies of the Proceedings or the announcement about IETF documents (and intermittent announcements when new Westine [Page 32] Internet Monthly Report OCTOBER 1987 documents are added to the directory) should send a note to Phill Gross (gross@gateway.mitre.org). Phill Gross (gross@gateway.mire.org) INTERNET MANAGEMENT Nothing to report this month. Vint Cerf (Cerf@A.ISI.EDU) PRIVACY Revisions were made to some areas of RFC989 (Privacy Enhancement for Internet Electronic Mail: Part I: Message Encipherment and Authentication Procedures), based on discussions and comments since the RFC was released. It is anticipated that an updated version of RFC989 will be produced following discussion at the Privacy Task Force meeting to be held on 4-5 November. John Linn (Linn@CCY.BBN.COM) ROBUSTNESS AND SURVIVABILITY No report received. SCIENTIFIC COMPUTING No report received. SECURITY No report received. Westine [Page 33] Internet Monthly Report OCTOBER 1987 TACTICAL INTERNET No report received. TESTING AND EVALUATION No report received. Westine [Page 34]