NLANR Measurement and Network Analysis Group (NLANR/MNA)
Passive Measurement and Analysis (PMA) Project~ Development of new metrics and real-time analysis for PMA One way to look at the Internet is to think about how you drive your car. You listen to the sound of the engine, the noise of the road, the feeling the steering wheel provides for grip on the road, the forces you experience in going around corners, the light conditions, rain, ice, etc. With the Internet, there are no such signals. The users, or network administrators, are flying blind: no sounds, no windows, no radar screen. The first time you notice is when something "doesn't work." The change is rather sudden, like the car falling apart in the middle of the highway at 70 miles an hour. No warnings whatsoever. The engine we seek to design should provide the kind of "noise" characterizing normal behavior, and put the user into the position to be able to detect abnormal behavior early, labeled neither good nor bad, just different. In time, this will be important. Preliminary efforts involved stepping out a method for developing a program in C to do real-time analysis and backporting Dag API support to the Dag3.2 cards. This was done in order to turn the SDA (SDSC Abilene) PMA machine into a test-bed real-time monitoring system. The backporting was successful and a 12 hour (night time) trace file was taken. Improvements were made to the network code including a command line option for users to access host for connection and porting. An application that takes data from the Dag card and grabs statistics from that record (e.g., average packet size, number of packets) was written. This early effort was followed by what would develop over the year into a four-prong approach; excellent progress was made in this work.
~ Special Traces In addition to our PMA Daily Traces Archive, we have data available from a number of special traces. These include a variety of OC48 trace data, an OC192 capture and a real-time analysis using smart, and long, contiguous traces. In addition to regularly formatted data, many traces have illustrated trace archives (analysis graphs), which help researchers to know about the trace before they download it. Also of use to our daily users, there are three index pages: alphabetical order,chronological order, and one with a subset of the data sets grouped for comparative studies. http://pma.nlanr.net/Special/ We have published 15 datasets so far, nearly half of which were added this year. Leipzig-I ~ An illustrated continuous 5-day GPS-synchronized IP header trace. http://pma.nlanr.net/Special/leip1.html Leipzig-II ~ An illustrated 1-day GPS-synchronized IP header trace captured simultaneously at either side of the University of Leipzig's central Internet access router. http://pma.nlanr.net/Special/leip2.html Changes were made to the anonymization tool to make it work for the GigE traces captured in Leipzig (thanks to Jesper Peterson of Endace for his help with this). NCAR-I ~ After NCG (the new NCAR Gigabit tap) was installed and collecting data, it was apparent that it is particularly busy, so the opportunity was taken to collect a long data set at the site, for one hour, 7GBytes, uncompressed. NCG is connected to a SPAN port and hence the timestamps might be distorted and some data could get lost on the switch before reaching the monitor. We began computing graphs for NCAR-1, however there is an old bug with xmgrace that prevents the PNG graphs from being produced and we are now into debugging this program, before the illustrated archive is available. http://pma.nlanr.net/Special/ncar1.html Auckland-VIII ~ The University of Auckland, is now instrumented with a DAG3.5E card. Data collected is between 100MB/hour (night) and 500MB/hour (day), in both directions. Post processing, copying, and archiving to the HPSS was a lengthy process.A surprise was that quite a bit of tool rework had to be done, as most of the graphing applications showed problems using the Endace ERF format. When the analysis of the Auckland-8 fifteen day trace was complete, we had over 3200 PNG graphs online, with many interesting details to be explored. A total of nine different graphs are produced, on an hourly and daily basis, and those are indexed in two different sets of HTML pages, once on a time basis (each set for one hour and one day, respectively) and another set per each graph, displaying a particular property along the time axis, with hourly and daily graphs. http://pma.nlanr.net/Special/auck8.html San Diego-I ~ A twelve (originally 30) hours IP header trace captured with an Endace DAG4.2GE dual Gigabit Ethernet network measurement card, at the end of January 2004. http://pma.nlanr.net/Special/sdsc1.html Tera-I ~ We collected data with the OC192MON-A at SDSC, connected to the Teragrid. Traffic is very light; activity increases for about 10 minutes every hour by an order of magnitude. First public 10GigE traces, collected with an NLANR PMA OC192MON located on SDSC's TeraGrid Cluster, began Sunday 8th of February 2004. http://pma.nlanr.net/Special/tera1.html Tera-II ~ 10Gigabit real-time analysis using SMART, a tool developed by Pere Barlet, running on one of NLANR's OC192MON's located on SDSC's TeraGrid Cluster. http://pma.nlanr.net/Special/tera2.html CESCA-I ~ Collected a GPS-synchronized IP header trace of the GbEth link that connects the Anella Cientifica (Catalan R&D network) to the global Internet. http://www.cesca.es/en/comunicacions/anella.html) A fairly busy Gigabit Ethernet link, this 34GB trace promises to be a very fine example which can be compared with the NCG and SDA measurement points. A highlight is the documentation of the CESCA-I data set, which is a very fine example on how much can be achieved in collaboration with other research partners: http://pma.nlanr.net/Special/cesc1.html. This page was done with quite a bit of help and background from Pere Barlet, who also got the final approval from the CESCA folks for publishing this trace data set. There is an illustrated trace archive for this data set. http://pma.nlanr.net/Special/cesc1.html Recently, the pma.nlanr.net server has been busy day and night computing all the graphs. We are waiting to be able to start the processing for the San Diego-I and Tera-I illustrated trace archives. ~ Development of OC192/10GigE measurements This has been a very complex process with many significant technical issues of various aspects that needed to be resolved. Our considerable efforts have lead to success in this endeavor. We presented a working 10GIGEMON in the SCinet NOC at SC2003 in November. This was followed by efforts to investigate and resolve problems with the payload content and to finding a configuration which works for a pair of cards in a system. Stephen Donnelly came to San Diego in December to work on the issues with running two Dag 6.1 cards in the same machine. With the latest release of the Dag6 software, and Stephen's considerable help, this problem was finally resolved. With the technical issues resolved, two OC192 monitors were then connected to SDSC's TeraGrid Cluster - two 10GigE lines going into the Juniper T640 Router and three load leveled Sonet lines being distributed out. We were able to connect to the two 10GigE lines, enabling us to monitor all the activity on that portion of the network. Data capture began on February 7, 2004 and continued until it was stopped on February 18. These first public 10GigE traces have been published on our Web site. http://pma.nlanr.net/Special/tera1.html We are ready to begin our planned OC192c instrumentation on the Abilene backbone. Placement details are now the primary goal. We have suggested two different scenarios for instrumenting Indianapolis (Abilene backbone). One plan would deploy just one OC192MON, the other is to instrument all three backbone links into IPLS, a router clamp with 2x OC192c and 1x OC48c monitors. We aliased the OC192-b machine at SDSC as TRG (TeraGrid) recently and it is collecting daily traces. ~ Additional new (and developing) strategically important passive measurement deployments This year we shipped our first PMA GigE measurement monitor (to the National Center for Atmospheric Research [NCAR] site). We previously had an OC3 monitor deployed there (NCAR has stopped using the OC3 connection.) In January 2004, we took a one hour IP header trace captured by a PMA monitor with an Endace DAG4.2GE dual Gigabit Ethernet network measurement card. The NCAR-I data set is published on our Web site: http://pma.nlanr.net/Special/ncar1.html The PMA team had multiple discussions regarding the future and evolution of the nai-p-sda (San Diego Abilene) monitor. Due to network changes at SDSC, we lost the connection to the Abilene network , so we pursued a connection to Abilene through the GigE CalREN (CENIC network) connection at SDSC, where we made use of the new regenerative tap. Development of the passive machine for GigE measurements was riddled with problems. The machine needed to be upgraded, as the new Dag4.3GE cards (GigE) required a PCI-X bus, usually a straightforward endeavor. However after a month long battle with our supplier, a bad backplane to the SATA controller, a bad motherboard, in general, a "lemon" which was returned to the manufacturer, to name a few of the difficulties, we persevered with the completion of the machine and installation on the regenerative tap. Data collection began and once stable, it was integrated into the 8x90 collection (daily traces). We expect some interesting results from this measurement point. Worked with Ohio State site; they are switching to a GigE connection, which they are interested in monitoring. We are arranging to create another GigE monitor for that site. Requests for passive (PMA) monitors were received from several strategically interesting sites. These new PMA deployments are in various states of preparation, construction, and installation. An OC12 monitor was shipped and installed at the Front Range GigaPop site in Colorado. It is currently collecting data. A request for a passive monitor was received from the KISTI site (Korean Institute of Science and Technology) in Taejon, Korea. A monitor was prepared and shipped; it has arrived and cleared customs. Currently there is no one on site to handle the installation and connection of the machine (our site contact has left to return to school). We are pursuing other avenues. A request for a passive monitor was also received from the AMPATH GigaPOP connection for the Florida International University (FIU) in Miami (nai-p-fiu). A monitor was prepared and shipped; it has been installed and connected. We are currently working with site personnel to resolve an optical connection problem. nai-p-i2a, Internet2 GigaPop in Ann Arbor, MI ~ An OC48 monitor was assembled, configured, and shipped to Internet2 in Ann Arbor, Michigan. The OC48 machine is on site (and reachable). We are working with Matt Zekauskas and Dan Pratt who are making the connections. nai-p-psc, Pittsburgh Supercomputer Center, Pittsburgh, PA ~ An OC48 monitor was assembled, configured, and shipped to PSC. Shipping was problematic with a delay of unknown cause, but the machine and splitter arrived in good condition. The CDMA timing unit was shipped shortly afterward. We are working with Kathy Benninger regarding this deployment. The machine is running; some connectivity, blocking, and security issues have arisen and are being addressed. The PMA machine at NCSA was disconnected when the OC3 connection went away. They now wish to connect it to an OC12 connection. We are planning to create an OC12 Dag3.2 monitor at that site. Towards the end of the period, we sent an introductory letter to Pacific North West GigaPOP, in an effort to perhaps be in a position to place multiple OC48MONs and lower speed systems in a prestigious location. We built, configured, and shipped a replacement machine for the University of Florida at Gainesville (nai-p-fla). The site is currently mounting the new monitor and requested additional mounting hardware, which we have supplied. We expect that they will in be in production soon.
~ Upgrades, troubleshooting, and maintenance on the existing PMA infrastructure A partial* total of 32 site problems in the existing PMA infrastructure (many sites had repeated issues) received attention this period. Five were still being investigated, or pending site action, at the end of the fiscal year. This was a productive time regarding PMA sites. A concerted effort was made to revive a number of OC3MONs. Working with site technicians, TXS, APN, and, OSU were brought back online and collecting data. PMA machines are often located in unattended sites. This makes it difficult to get assistance with the machine to make corrections. Site support people must travel from their normal locations to the unattended site. In one case, our support contact traveled approximately 100 miles. To continue the success with reviving existing machines and bringing problematic new deployments online, plans were made for Jörg Micheel, PMA project lead, to visit the remaining sites and work with their personnel in April and May. A survey of the existing gear towards the end of the fiscal year, gave the following insight: OC3/OC12MONs cards for 3 systems available (1 system ready to ship) We also have a stock of UoWaikato DAG3.2 cards left, and we are considering to not use them except if unavoidable, as these are now three years old and are out of warranty and support and would not do any of the modern works we are interested in, such as real-time analysis. We are short on any type of splitters, if requests would come in. A test bench for the Dag cards and PMA monitors was created. We worked with Kevin Walsh (SDSC NetOps) who helped configured the test equipment. This will ensure that all Dag cards are shipped in working condition. Grant Duval of CAIDA offered to work with us using the Smart Bits equipment for testing PMA monitors. Due to progressively worse problems with the PMA server, we made plans and did the groundwork regarding replacing/upgrading the server. It was nearly five years old; the last downtime for the machine was December 2001. A new PMA server was prepared (fitted with an 8 port 3ware parallel ATA Raid Controller with 4 120GB IDE Hitachi/IBM Hard Drives). In October a new PMA server was installed and migration of data completed. HPSS was used extensively in the process, including recovering some files that had been previously lost, and improving robustness. In addition to other installations and configurations, we did another rsync from the /traces directory on the old machine to the new, which puts the new subscriber system in place for the Trace User Community. We added the additional disk chassis giving the server near a terabyte of storage, thereby doubling the available disk space for traces. We restored all the online traces on the pma.nlanr.net server after the upgrade to 1TB of storage space. All of the long trace files from the various sources are now online and represent roughly 60% of all the data accessible. We also restored most of the meta data and the collection of old trace files (samples of every 15th of each month since the beginning of collection). The 1TB will be filled up, which forces us to plan well ahead for the next upgrade around the beginning of the next fiscal year, since we intend to bring online more long trace files. We increased the RAM to two gigabytes. This is needed to support trace analysis processing on the machine. Of interest is that problems with the PMA server were first noticed by a number of users who were keen to download more OC48c traces. For those with desperate needs, the data was made available in a different location. It was very refreshing to see how many people really *needed* the trace files, as a matter of urgency. During a period of four days, five researchers were asking for OC48c traces and long traces to be made available quickly, as they were working on a paper or contribution with a deadline. Auckland-6 and a portion of Abilene-I were mirrored for folks. Inquiries came from UNC, USC, UMASS, UWISC, and from universities in the UK and Brazil. We are looking at the possibility of using a CDMA modem such as the AirLink Redwing as a timing source to time stamp PMA traces. If this can be made work it will reduce the cost of time stamping considerably over the use of the EndRun Technology CDMA device. Serious issues regarding damage done during shipment of the passive monitors arose this year. We have been gluing parts into place as well as adding internal packing to the machines. However, more than one machine arrived at its destination in poor condition, causing installation problems and the need to send replacement parts. This also inconveniences the site personnel to a large degree. Therefore, we pursued alternate shipping methods. We are working with a professional shipper to ship PMA machines now. We have had a special shipping container and packing constructed, which can be reused for multiple shipments and will provide the protection we need in our shipping. The cost for this was quite reasonable and we think it will prove quite valuable. More detail on these activities can be found in the monthly and quarterly reports for this reporting year (April 2003 - March 2004), available at: http://moat.nlanr.net/Reports/
Active Measurement Project (AMP)~ Reimplementation of AMP and the development of a new testing architecture Excellent progress was made on this complex coding project. At the end of the fiscal year, the beta release for the AMPlet package (the programs located on the site monitors) was ready to be released. Web pages have been prepared. We are just waiting on details of the appropriate public licence under which to release it. Code has also been written for AMP throughput tests and pathchar (a bandwidth estimation tool) and refined for integration into the main distribution. Therefore, we anticipate that soon we will release a second beta version of the AMPlet package; this one with scheduled throughput tests. Before beginning the coding of the AMPlet package, an architecture for measured (the measurement daemon) was drawn up and plans for the implementation were begun. During this work a fair amount of effort was devoted to setting up the infrastructure for the whole of the reimplementation project. Important features of the new AMPlet package:
While coding IPMP we wanted to check the checksum we were generating from a tcpdump trace and went looking for a tool that we could feed a string of hex from tcpdump into which would give us the checksum. We could not find one, so we wrote one in Javascript; it is something we have often wanted to do, so we are guessing it will be helpful to others as well. http://moat.nlanr.net/Software/HEC/index.html ~ IP Measurement Protocol (IPMP) Version 04 of the IPMP IETF Internet Draft was submitted to the IETF RFC editor. Comments received from Randy Presuhn and others were addressed in this version. Mark Allman has setup a "review team" for reviewing the IPMP protocol to take it forward within the IETF. http://www.ietf.org/internet-drafts/draft-mcgregor-ipmp-04.txt AMP Project lead, Tony McGregor, met with the CEO of Allied Telesyn in NZ (they do the development work for their layer 3 devices in NZ). He has agreed to revisit the idea of them implementing IPMP. Development of IPMP continued with work on ipmp_pathchar. It is an ncurses application that works to estimate the serialization rate of each link on the forward and reverse paths (shows a list of segments and the speed in kbps that it sees, reporting results in "real time"). Additional work was performed on the Linux and FreeBSD kernel implementation and the addition of the ability to specify the 8 TOS bits in the IP header for IPMP echo packets, which at least one AMP site might find useful. State diagram figures for IPMP for describing how the protocol works were created. Measurements to profile the behavior of CRCnet's quickbridge wireless link were taken. There is an odd hockey stick effect in the RTT that is caused entirely by the forward path. The reverse path is predictable and constant (perfect for capacity estimation), the forward path less so. http://www.wand.net.nz/~mjl12/quickbridge.gif The ability to identify the bottleneck capacity and the position in a path where a bottleneck occurs is helpful for network operators in understanding the performance seen on a path. Current capacity estimation techniques measure network paths with varying levels of accuracy, speed, and robustness. Tools like pathrate have become progressively more sophisticated and accurate at measuring the capacity of a tight link, but do not provide any indication of where the tight link is located in the path. pathchar-like techniques attempt to measure the capacity of each hop by exploiting the Time To Live (TTL) value, but it is difficult to isolate the delay contributed by a specific link when a noisy link precedes the link measured. We have developed a technique to segment a path into isolated links for capacity estimation purposes, by time stamping the probe packets at the ingress interface of each router in the path. The technique uses IPMP to obtain timestamps at each point in the network, but could use any other tracing protocol that provides a timestamp of sufficient resolution. The technique can be used to estimate the capacity of each hop by a method that isolates each hop in the path in an end-to-end measurement. This technique allows for forward and reverse path measurement and does not require the timestamp sources at each router to be synchronized. IPMP Bandwidth estimation experiments and result analysis were performed. A basic finding to date is that if the second packet is larger than the first, then the first will get further and further in front of the second the more store-and-forward devices (routers / switches) it traverses. This is because the first is received and transmitted in full before the second packet is received in its entirety.
A program was written to generate cross traffic (CT) based on packet sizes and packet inter-arrival times seen in a particular trace. It does not replay the trace. The point of this is to have a cross-traffic-from-trace generator/application that creates traffic based on that seen in the real world on the emulation network. The first application of this was CT experiments using the IPMP bandwidth estimation techniques. http://www.wand.net.nz/~mjl12/emulation/ Experiments on the WAND emulation network using a 10mbit link in the path immediately after a 100mbit link, using the IPMP packet tailgating method and the ctft generator were performed. The data is used to empirically show the difficulties of measuring the capacity of a link that is much faster than the link that fed it (e.g., going from a 10mbps link to a 100mbps link). Results were submitted to PAM2004: Segmentation of Internet Paths for Capacity Estimation. http://www.wand.net.nz/~mjl12/pam2004.pdf ~ IPv6 and IPv6 Scamper The AMP IPv6 mesh continued to grow. One of the new international AMP sites, amp-surf (SURFnet GigaPop in Amsterdam, Holland), was among those who requested that they be added. There are now 13 sites in the AMP IPv6 mesh. http://amp.nlanr.net//IPv6/ A series of graphs that plot delay statistics for the 8 months of both IPv6 and IPv4 was produced. http://mave.nlanr.net/~mjl/ipv6-longterm/I Analysis was performed and graphs created as a case study for the IPv6 handout. http://watt.nlanr.net/IPv6/ Some quite substantial performance increases at the University of Oregon were observed using the IPv6 mesh. An IPv6 route change can be seen at approximately 10pm 13 May 2003 PDT. We have written a topology mapping tool, IPv6 Scamper, to traceroute through a large address list in an efficient manner by conducting traceroutes in parallel - as the need dictates - to fill a packets-per-second rate. Scamper is quite similar to CAIDA's Skitter, except that Scamper can traceroute to IPv6 addresses, as well as IPv4 addresses. The focus of the current research is in providing insight into the behavior and growth patterns of the IPv6 Internet. The goal is to compose an address list that is comprehensive in that it will have an IP address from each prefix. We have composed an address list, which has been Scampered from several locations at various times. sorcerer.cs.waikato.ac.nz, via freenet6, Scamper was run on all IPv6 AMP machines from amp-nysernet. (Scamper traceroutes in a parallel method to a list of IPv4 and IPv6 addresses.) A tool was written such that, given a list of IPv4 / IPv6 addresses, it identifies which addresses belong to the same router (tool is named "groupie"). The plan is to take multiple Scamper views of Abilene's IPv6 network, run the output through groupie, and then generate topographical maps with graphviz dynamically (as is done with the other AMP visualizations). http://www.wand.net.nz/~mjl12/sorcerer-nysernet-wide.png Code adding the ability for Scamper to detect if a path is dead after receiving no responses for N hops is close to implementation. Additional functionality was added to Scamper to enable it to continuously traceroute, and to detect when the address list changes. Support to stop tracing to an address if it detects a loop was also added. Efforts to profile the length of loops discovered during the Scamper IPv6 runs took place. This is a relatively important statistic given that 25% of all addresses in the list fail with a loop detected. http://www.wand.net.nz/~mjl12/ipv6-scamper/ A C API was developed for accessing Scamper output files - both the ASCII text style and a binary file format previously designed. A series of programs to perform rudimentary analysis on the data collected from the six Scamper runs that have been taken from various points in the IPv6 Internet were written. http://www.wand.net.nz/~mjl12/ipv6-scamper/ In response to a community need for tools that collect statistics on paths that support a >1500 byte MTU, we decided to implement a path MTU discovery function (PMTU support) in Scamper; for the purpose of finding tunnels. An IPv6 path-MTU discovery was conducted between all the AMP IPv6 monitors; It successfully found the existence of a few tunnels. The AMP mesh was used to show that the anycast address for routing to the closest 6to4 gateway is not well done in the US (or there are no good 6to4 gateways in the USA). Basically no AMP machines on networks were found that route to a 6to4 gateway in North America. Also it was found that most of the international AMPs were routed to close-by 6to4 tunnel entry points. A new program was written that picks out alternate paths out of Scamper traces. The idea is that short IPv6 routes may use tunnels over IPv4, which may not be the most direct route and lead to a higher RTT despite being a shorter [AS?] path. We wrote an application called `ring', which sends IPv6 strict-source routed packets to measure the round-trip over a specific route. Using the last Scamper run produced by William F. Maton (National Research Council, Canada), we found 178 alternate paths. A few were probed with the 'ring' tool to check that it works. Maton has been collecting Scamper runs for about 4 or 5 months now from one source to a constant target list. He has generated data that will enable us to compare IPv6 routes that have changed over time. We have performed some analysis on the weekly Scamper traces that he is generating for us. http://www.wand.net.nz/~mjl12/ipv6-ryouko/ ~ New Path Display tool - divide and conquer graphic (Path Viz) Work began on this new tool with which the program's graphic output will display the path under investigation, and show portions of the paths it shares with nearby monitors to help locate problem paths. ~ New (and developing) strategically important active measurement deployments We set up an amp-naukanetnwu to test to a set of machines (30-40) in Russia, to which Greg Cole (NaukaNet) arranged for us to test. In addition to providing interesting data, this will also provide a stepping stone to a Russian AMP mesh. The monitor point on the NaukaNet network is Northwestern University in Chicago, Ill. Working with Rick Summerhill and other folks from Internet2 about putting AMP monitors into Internet2 nodes. The goal is to locate AMPs at the 11 Abilene backbone nodes as part of the Observatory Project. The first of these machines has shipped (with the new motherboard discussed below). amp-i2at (Internet2, Atlanta) is now up and collecting data. The following are interesting new sites brought online and collecting data this year in the AMP network:
Helsinki U. of Tech. (HUT network in Finland) is the first site to be in the "international only" mesh (HEANet in Ireland will be the second). The placement of AMP monitors at GigaPops is significant because it provides for research such that traffic performance over backbone segments between GigaPOPs can be studied versus end-to-end performance between hosts using those same segments. The CNIC group in China completed the AMP request form recently. The TransPAC connection is moving from the Pacific Northwest (PNW) GigaPop in Seattle to a location in the Los Angeles area. Our AMP monitor at the PNW GigaPop was on the TransPAC connection, therefore it must be moved, or associated with another network. We worked with the site The AMP monitor placed at the Pacific Northwest GigaPop in Seattle by TransPAC is in return shipment. PNGS people are willing to host the monitor but need a source of funding for the minimal rack space cost. The plan regarding the new location for the TransPAC connection (Los Angeles area) is to communicate with John Hicks and install an additional AMPlet in Los Angeles to cover the TransPAC connection. We heard from Brian Court, of CENIC, that he would like to have AMP monitors at seven CENIC sites in California (each of the hops). 10GigE AMP ~ We began research to gather information in support of the assembly of an 10GigE AMPlet. We are interested in taking active measurements at this speed. CISCO has expressed interest in this. Les Cotrell (SLAC) and Phil Dykstra (WareOnEarth) are interested and engaged in pursuing this as well and have been very helpful. ~ Upgrades, troubleshooting, and maintenance on the existing AMP infrastructure A partial* total of 79 site problems in the existing AMP infrastructure (many sites had repeated problems) received attention this period. Three were still being investigated, or pending site action, at the end of the fiscal year. In the Fall, we had a spike in the number of problem AMP machines which turned out to be du to sites blocking ICMP traffic. This coincided with the blaster worm. Apparently, the nachi worm was created as a counter to the blaster worm. But it seems the nachi worm caused router tables to fill and create DOS conditions. Some AMP machines were turned off as a result of the chaos. At its peak fourteen AMP machines were either turned off or had ICMP blocks at the routers. Our efforts working with site personnel were successful and AMP site outage returned to the levels enjoyed before the rash of worms. We continue to monitor for ICMP blocking and contact sites as it arises. Even though the newly discovered vulnerability on OpenSSH did not represent a threat to AMPlets at remote AMP sites, we decided to move ahead with the update to the 3.7 version to ensure that remote site people do not perceive the monitors as vulnerable. Data collection and transfer were not affected by this. Computer component obsolescence is an issue with which we are continuously faced. The component that becomes obsolete and out of manufacture most is the system board (mother board). It has become necessary again to test and upgrade to new system boards as the current board (GigaByte 6VML board) will go out of manufacture and is becoming unavailable. Investigation revealed that only a few manufacturers are continuing to produce boards supporting the Pentium III processors. Boards supporting Pentium IVs are much more available now. This necessitated a switch of processors to the P-4. After extensive testing, we chose a new board; at which point our vendor notified us that the board has been discontinued. We found an alternate supplier, retested and chose a new board. The only drawback is that AMPlets on gigabit interfaces will require add-in NICs as in the past. We were finally able to order the new AMP machines (the first of which are destined for the 11 Abilene backbone nodes as part of the Observatory Project). AMP staff discussed the implementation of the serial console on AMPlets. We worked to research the implementation of the serial console and it developed to be quite straightforward in FreeBSD. The kernel we ship already can do it, we just have to make small adjustments to two files (/boot.config and /etc/ttys). This may be something to be deployed on all AMP sites since it allows for the use of a laptop instead of a keyboard and monitor to examine misbehaving AMP monitors. This capability had been requested by site amp-surf (SURFnet, in Amsterdam), it has now been implemented on the SURFnet monitor. The archiving process performed on the AMP data collection servers (AMP and VOLT) moves the active measurement data older that six months to the HPSS (SDSC's high performance storage system). When this method was started some time ago, this process would take the data disk fill status down to approximately 68 percent. However the active measurement infrastructure has continued to grow. The archive process now takes the disk fill down to the low to mid eighties percentage. This indicates that with the growth of the AMP meshes, the accumulation of data over a six month period has grown considerably. In addition to the disk fill being reduced less, archiving is required more frequently now (approximately every other week vs. the previous four to six weeks time frame). It is important to note that the am_slave process failure occurs at about ninety-one percent fill. For the first time in a couple of years, we had a disk crash in the servers. Plans and specifications to upgrade the servers were developed, researched, and put into place. Use of a RAID controller would remove many hours of work when we lose a drive and ease recovery from faulty disks. New drives could hugely increase the capacity of the array and lengthen the time between backup runs. There was some improvement in the functioning of the AMP and VOLT servers with the addition of the new 36GB drive installation. The disk fill situation emphasized the need to move a quickly as possible on the new replacement servers we are acquiring for AMP and VOLT. New machines were purchased in January; preparation of the machines and what tests will be performed were discussed. It was decided that we should determine if the new servers will be FreeBSD or Debian Linux. Therefore, a basic FreeBSD 4.9 operating system was installed on one of the new AMP/VOLT servers and Debian Linux 3.0.23 (woody) on the other new server. Initial performance tests between the two new amp servers showed the FreeBSD one is beating the Debian one well and truly. However, the Debian one is running on a 2.2.20 kernel so we have been trying to build a new kernel for it. As the NLANR/MNA infrastructure grows, a need for better access to remote monitors has been identified. We continue to research options. For some time now we have looked at simple remote rebooting methods such as the DataProbe iBoot and the CPS Heart Beat Re-booter. A short time back we implemented a remote serial port console access to the amp-surf (SURFnet in Amsterdam, Holland) monitor. That is proving very valuable to diagnosis anomalies. However it requires access through another host on site. Other options are the Intel IPMI chipset and the Cyclades Corp., PCI bus Remote Access AlterPath SM100 card solution. We are testing both an IPMI and a Cyclades SM100 card. (This is for both AMP and PMA needs.) Efforts moved forward with regard to hiring the AMP Software Engineer, who's primary task will be to work on the reimplementation of the code for the AMP data collection servers. 16 applicants were reviewed. 10 have taken the programming skills test, and three have been interviewed in person. References are in progress on the two top candidates. More detail on these activities can be found in the monthly and quarterly reports for this reporting year (April 2003 - March 2004), available at: http://moat.nlanr.net/Reports/
Wireless Performance Measurement and Analysis ActivitiesWork was performed on the HPWREN utilization reports that summarize the usage of HPWREN at several key locations - including the Palomar and Mount Laguna Observatories, the Santa Margarita Ecological Reserve, and the Tribal Digital Village Network. These reports focus on how each location uses HPWREN to its benefit and how effective HPWREN has been in providing high-speed internet access to areas where it would otherwise be unavailable. Each report also traces the use of the network from the time it was first installed up through this pas August, and demonstrates to what extent usage of HPWREN has increased in that time span. http://stat.hpwren.ucsd.edu/HPWREN/Test/Reports This work will provide a baseline and comparison point by which we can measure the changes in network utilization from the beginning of HPWREN to the present. This will hopefully lead to an understanding of how small user communities change in utilization after the network goes from new to truly useful. The measurement infrastructure of HPWREN (High Performance Wireless Research and Education Network) was further developed. Tools to visualize network measurements with graphs were written and existing tools and Web pages significantly improved. The Stratex Radio graphs: added an option that lets the user view data in two directions at once on the same set of axes. http://stat.hpwren.ucsd.edu/HPWREN/New/Stratex/stratex.cgi Overhaul of the Lucent and Tsunami MIB page: to make the page easier to view and use, it was split two separate pages with an easy to use index at the top of each page. Another big improvement is that new sites are detected and displayed automatically. http://stat.hpwren.ucsd.edu/HPWREN/New/Lucent_and_Tsunami_MIB_page.cgi http://stat.hpwren.ucsd.edu/HPWREN/New/Lucent_MIB_page.cgi Improvements to the throughput graphs: This page not only allows throughput measurements to be graphed all across the network, it also displays the most recent throughput measurement. http://stat.hpwren.ucsd.edu/HPWREN/New/thruput.cgi Two ongoing projects: the Snort intrusion detection website, and the IQEYE camera. The Snort Web site displays graphs of the rates at which certain IP addresses send illegal packets across the network and graphs of the total amount of intrusions detected over user-specified time periods. There is also a network intrusion status page and a daily email report in development. A device driver has been designed and implemented for an IQeye wireless camera. http://www.iqeye.com/prd/IQe3W.htm The camera is planned to be mounted on the back of a plane to take pictures of the ground in flight. It will also be connected via serial port to a GPS device from which it will correlate GPS coordinates with pictures taken. The pictures will be stored on camera until the plane flies within range of a wireless access point on a mountain top where they will be transmitted to a server. A script was written that sends out email alerts when UPS devices lose power or enter an error state. Some network analysis based on HPWREN packet traces was performed. For an initial 19:49 to 06:20 run, Palomar was clearly dominating the traffic: bytes 33991205509 Palomar Observatory The total in the list amounts to 36595974564 bytes (which is not much more than the Palomar traffic alone), though Palomar had stated that they would like to do 100GB/day or more eventually. The duration appeared to be about (10*3600)+((20+11)*60)=37860 seconds. 36595974564bytes/37860seconds=966613Bps. 966613*8bits=7732904bps, or almost 8Mbps, on average. More analysis was performed over a weekend, during transfers from Palomar to Yale, the links from the Observatory to SDSC are running pretty much flat out. The traffic analysis work led us to start on a pcap-based data collector we intend to run on a dedicated trace analysis machine that should continuously collect data. More detail on these activities can be found on the HPWREN Web site: http://stat.hpwren.ucsd.edu/
Outreach, Utilization Improvement, and Documentation Activities~ Impact: Data Users and Citings We have begun to gather and compile information on our data users and how they use our data and tools. The Citings: Data Users Web pages are a part of this effort. The pages have information on publications, researchers, and educators referencing our work. We currently have pages on published papers referencing our work, sorted by meeting and/or conference, sorted by year, and cumulative numerical information regarding proceedings where we are cited. Example: Of 29 total accepted papers at IMW 2001 (San Francisco, CA, USA November 1-2, 2001), 11 papers referenced our work (37.9% of the total number of papers accepted). http://mna.nlanr.net/Data-users/ Following are some interesting comments and user experiences from this past year: Problems with the PMA server were first noticed by a number of users who were keen to download more OC48c traces. For those with desperate needs, the data was made available in a different location. It was very refreshing to see how many people really *needed* the trace files, as a matter of urgency. During a period of four days, five researchers were asking for OC48c traces and long traces to be made available quickly, as they were working on a paper or contribution with a deadline. Auckland-6 and a portion of Abilene-I were mirrored for folks. Inquiries came from UNC, USC, UMASS, UWISC, and from universities in the UK and Brazil. While working on the IPv6 project, we found a couple of IPv6 routing anomalies out of NYSERnet. We reported them to Bill Owens (a longtime collaborator, systems engineer for NYSERnet), who was then able to correct them. One of their routers was not taking routes from a particular source which made chunks of their IPv6 space non-routable. We did research on the impact of NLANR/MNA work and discovered that a total of eight accepted (and presented) papers at the ACM SIGCOMM Internet Measurement Conference 2003 in Miami, FL, used MNA data for their research and analysis works. The following paper briefly discusses IPMP in the context of network troubleshooting from the user's computer. Sections 3.2.2 and 6.2 are the interesting things for IPMP, but the whole paper is worth a read. User-level Internet Path Diagnosis, Ratul Mahajan, Neil Spring, David Wetherall, and Thomas Anderson SOSP, October 2003. http://www.cs.washington.edu/research/networking/tulip/bits/sosp2003-tulip.pdf ~ Papers, Presentations, and Conference/Meeting Participation Micheel, Jörg. Writing Applications for Real-time Network Monitoring. Proceedings of the AUUG 2003 Conference: Open Standards, Open Source, Open Computing, Sydney, Australia, Aug.-Sep. 2003. McGregor, Anthony J., Perry Lorier, and Mark Hall. Flow Clustering Using the EM Machine Learning Algorithm. Accepted by PAM2004. Mochalski, Klaus and Jörg Micheel. A Real-Time Packet Burst Metric. has been accepted for the Terena Networking Conference 2004 in Rhodos, Greece, in June. Luckie, Matthew and Anthony J. McGregor. On Segmentation of Internet Paths for Bandwidth Estimation. Submitted to the SIGCOMM Internet Measurement Conference (IMC) 2003. Lee, Jae-Min, Jian-Bo Gao, Ronn Ritke, and Tony McGregor. Characterization of end-to-end packet dynamics in the Internet. Submitted to Sigmetrics 2004. Another version of this paper will be submitted to the Special issue of Performance Evaluation. The following abstracts were submitted to PAM2004: Gross, Christopher W., Jörg B. Micheel, and Hans-Werner Braun. Design and Implementation of a Scalable Real-Time Network Sensor. Luckie, Matthew and Tony McGregor. Segmentation of Internet Paths for Capacity Estimation. McGregor, Anthony J., Perry Lorier, and Mark Hall. Flow Clustering Using the EM Machine Learning Algorithm. Accepted by PAM2004. Mochalski, Klaus and Jörg Micheel. A Real-Time Packet Burst Metric. Gannis, Mike and Ronn Ritke. NLANR/MNA Builds an International Infrastructure for Network Research. [White Paper] When first released SDSC made this the highlighted news article of their front page and is now available in the news section at: http://www.sdsc.edu/Press/features/120203_NLANR.html ~~ By invitation, Ronn Ritke and Tony McGregor attended the RELARN meeting in St. Petersburg, Russia. Greg Cole (NaukaNet) attended as well. Both Ronn and Tony gave presentations on our work. In addition to giving a presentation, Tony gave a demonstration of AMP (from his laptop, as there was no Internet connectivity). The conference went well and we seem to have built some real bridges with a commitment to host an AMP monitor in Russia and to include measurement in the GLORIAD network. Greg Cole (NaukaNet) was very pleased with how the conference went. From an email Greg Cole sent to the U.S. participants: Matthew Luckie attended the IETF57 meeting, joining several IPv6 sessions, as well as the one regarding IPMP. Jon Bennett, who has developed a strong interest in our IPMP protocol, had discussion moved from the IETF IPPM working group to the TSVWG working group. Matthew attended the session and spoke from our perspective, which differs significantly from Jon's at some points. Bud Hale attended an afternoon session of the North American Global IPv6 Summit held in San Diego at San Diego State U (SDSU). Besides the presentations, he enjoyed the opportunity of meeting some of the people attending from locations with AMP and PMA monitors with whom he often works. Jörg Micheel attended NANOG28 in Salt Lake City. Netforum 2003 conference (NZNOG 2003) Jörg gave a 60 minute presentation about the state of the art in passive Internet measurements. Attended the other sessions and was positively impressed by the quality of the presentations. Had the impression that the issues raised were much closer to people's needs on a day-to-day basis. Tony presented on the simulation work we did a year or two back which created more interest than expected. Matthew gave a talk about IPMP and the IPv6 mapping project and attended. Created a large poster with the IPv6 map for NZNOG 2003. Produced a map consisting of 56 pages of A4 with flags of the major points on each AS. It attracted some attention, Joe Abley who is an ex-pat NZer now working at ISI offered a box to run Scamper from. As a result of the NZNOG conference we have offers to host about 10 AMP monitors in NZ so that will make the core of a national mesh here. Jörg presented his paper, "Writing Applications for Real-time Network Monitoring" at the AUUG2003 Conference (Sydney, Australia, September 3rd-5th). Jörg attended ACM SIGCOMM 2003 in Karlsruhe. It was not only a data communications festival, but a prime opportunity to talk to a lot of people in the network arena. Jörg attended the Internet2 Security at Line Speed (SALS) NSF Sponsored Workshop (Invitation Only, total of 30 attendees), in Chicago, IL, August 12-13, 2003. The outcome was not unexpected, but educational. Most people will treat security as a matter of restricting IP communications by moving "wide open" to "tailored to needs" via firewalls and similar "middleboxen". More interesting was the fact that tools are all there, no surprises expected in the near future, but changes are expected in the way that those mechanisms are operated. Multi-tiered policies, better ways to communicate, multi-campus arrangements and similar communications and cultural changes promise some form of "improvement" on the subject. Overall, reassuring to see that there are no strict technology changes expected. http://apps.internet2.edu/sals/2003aug/agenda.html Tony visited Canterbury and Massey Universities presenting material about what NLANR/MNA is currently doing, in the hope of finding more collaborators. He also attended a talk about NeTraMet++, given by Nevil Brownlee. Tony attended a seminar given by a visitor from Cardiff University about network metrics and QoS for grid computing which raised some interesting issues. Ronn attended the Kansas Joint Techs TEV meeting in Lawrence Kansas; participated in the PIPES meeting; and attended the JET meeting. Ronn participated in the SDSC Measurement Working Group meeting. There are plans to have some kind of measurement meeting at SDSC in December. Lana Kennedy and Ben Reesman each prepared and presented a poster at the SDSC sponsored Student Symposium for student researchers on their current work and activities. Joint Engineering Team (JET) Meeting, Arlington, VA, Oct. 2003: Tony gave presentation on NLANR AMP to the JET meeting. Ronn introduced Tony's teleconference presentation, and attended most of the meeting and met with a number of people including Rick Summerhill. Tony gave the NLANR AMP talk at Auckland University; this was the last of the NZ talks. Internet Measurement Conference (IMC),Miami, FLA, Oct. 10-27,2003: SC2003, Phoenix, AZ, Nov. 15-21: WIDE-CAIDA-RIPE-NLNET workshop at ISI, Nov. 7, 2003,Marina del Rey, CA: INTERNET ENGINEERING TASK FORCE(IETF), Minneapolis, MN, Nov. 9-14 2003: Jörg had visitors, a Professor from Finland and students from Finland and Sweden, and some from University of Waikato. He gave them a 90 minute overview presentation on passive measurement research and technology. Internet Statistics and Metrics Analysis (ISMA) Bandwidth Estimation Workshop, San Diego, CA, Dec. 9-10, 2003 Measurement Architecture Workshop, SDSC, CA, December 11-12, 2003 Attended the Town Hall Meeting at SDSC. People from NCSA visited SDSC to work on TeraGrid plans. Ronn Ritke participated in one of the breakout sessions that had a network focus, and met with Tony Rimovski to go over presentation slides for Feb. Ronn Ritke attended the Joint Techs/APAN meeting. He gave a presentation on NLANR/MNA as part of a measurement session. Completed three sets of presentation slides (research, project overview, Joint techs/APAN) and a one-page handout for the meeting. NSF PI meeting in Arlington, Virginia. Ronn Ritke attended and had meetings with John Towns, Peter Arzberger, Fran Berman, Matt Mathis, Bill Chang, and others. He also gave a poster presentation Thursday evening and a slide presentation on Friday. Ronn Ritke attended the CENIC Meeting in Marina Del Rey CA (Mar.2004). Chris Thomas from UCLA will install a new Gig interface for the UCLA AMP machine. Ronn met with Greg Hidley and discussed the OptiPuter project. Ronn also distributed handouts on the 10GigE PMA traces and copies of the NATimes. Quarterly Student Presentations: ~ Documentation, Networked Data, Publications "NLANR Holds 'Very Successful' Workshop on Passive and Active Network Measurement" was published in Online, Vol 7 (8), April 16, 2003. Online is the SDSC/ NPACI biweekly newsletter. http://www.npaci.edu/online/v7.8/pam2003.html Proceedings of PAM2003: the Passive and Active Measurement Workshop, La Jolla, CA, April 2003. Published a new issue of the Network Analysis Times, the theme of which is our AMP IPv6 efforts. http://moat.nlanr.net/NATimes/NAT.4.1.pdf The Network Analysis Times was distributed at the North American SC2003 Global IPv6 Summit held at SDSU, at NZNOG, and other meetings. Distribution of the most recent issue continues throughout SDSC. A handout about our AMP IPv6 activities was created for distribution at the North American Global IPv6 Summit held in San Diego at SDSU. http://watt.nlanr.net/IPv6/June03Handout.pdf An International handout was created regarding our international efforts and deployments. The AMP, PMA, and general MNA posters were updated (and displayed at PAM2003 as well as over venues). Several extensive Web projects involving the development of many new pages, updating of current ones, and in some cases, realignment of the underlying tree structure, were begun. A number of significant changes took effect and continue to be in progress for the Web pages. The primary goal is to increase readability, accessibility, update content, and for the home page particularly, develop living, dynamic content. The Web page layout/style was redesigned, including the creation of a new navbar (to include rotating images which facilitate and draw the viewers attention to changed content). The style and design is consistent across the new pages (and will be implemented on current pages as time allows). There are different navbars, depending upon whether the page is a general MNA page, or an AMP, or PMA page. The underlying design of the navbar is such that these featured activities can be changed at any time without having to change the html on all pages. The plan is to rotate them every few weeks or so, and of course add new ones as we highlight different activities. All design elements were checked for cross-browser compatibility; embedded and in-line style sheets are used to accommodate many of our users who use older browsers. Templates were created for MNA, PMA, and AMP pages. The nature of our pages which contain both text and illustrated data of varying sizes (some quite large) necessitated development of two page templates: one for regular pages with images that fit easily into the template size constraints and another for "data" pages, such that the researcher is not confined to a specific size when displaying illustrated data (graphs and/or images), while maintaining continuity and readability. A mini version of the navbar (just two links vs. the full sidebar one) for use with the data pages (where the complete sidebar version is not appropriate) was designed. We designed and created an m4/makefile system for making changes to the Web page templates, not just format/style changes, but significant changes in the navbar, for example. This approach allows for the creation of Web pages that use the templates without replicating the template within every page. CSS was not the best answer for our Web pages for two reasons. First many of our core users do not use browsers capable of properly rendering the latest CSS functions. Second, it applies primarily to style elements and not structural aspects of the page. A new home page has been designed, developed, and implemented. Two elements of the page are intended for dynamic content, the "Latest Pings" which will have brief overviews of our most recent featured activities and a "Highlights" column with quick links to our most active efforts. http://mna.nlanr.net Citings: Data Users pages ~ The aim of these pages is to know who is doing what with our data, and approximately how many folks are using our publicly available measurements, analyses, and tools, for their own research efforts. The Citings/Data users pages currently consist of the following pages: Published papers referencing our work, sorted by meeting/conference and sorted by year and a "By the Numbers" page with numerical information regarding proceedings' papers in which we are cited. The results of this work which spanned many weeks of compilation efforts form an excellent foundation which will be extended as new information is added. The pages are posted at: http://moat.nlanr.net/Data-users/ Work began on the redesign and restructuring of all of the PMA Web pages (including directory tree structure). Nine primary pages that will form the core of the new PMA Web site were identified. The PMA navbar was developed based on these primary pages. A new PMA site map where the site markers are color-coded per the speed of the link measured was created. http://pma.nlanr.net Developed three index pages for the Special Traces: alphabetical, chronological, and one where a subset of the traces are grouped for comparative studies. http://pma.nlanr.net/Special Work to develop a Web logs system took place; we anticipate having the statistics for ftp downloads on Web pages soon. A redesign of the AMP map of remote sites (for both static publication and use on the Web pages) was implemented. Population of the map is based on latitudes and longitudes of site locations and is an overlay of a static topographical map. There are international and US only versions of the AMP splash page maps, with small and large versions of each. http://watt.nlanr.net/active/maps/ampmap_active.php Based on the Internet2 AMI service, Warren Matthews is making good progress creating an AMP Web Services implementation for us. (Thanks Warren!) http://watt.nlanr.net/active/ampwebservices.html A new flag (status) was added to the AMP database which indicates that a monitor is no longer active. Setting status to disabled means that no tests are done to the monitor, but the data is still available in the Web pages. Some user initiated additions were made to the AMP throughput page. Discussion is still ongoing regarding the mechanism we should use for preserving state for the new AMP Web pages. We want something that works well over a range of different user types, in particular users who do not use cookies. We did some work on the Otter topology display, adding options for displaying just the most common path each day and for omitting the tabular display. We also updated to the most recent otter. We hit a bug in otter, but Brad Huffaker (from CAIDA), who is the main author of Otter, was able to fix it for me and do a new release. This version prints even sized nodes, which is good for our use of it. Installed the HEC calculators on the MNA software page. http://moat.nlanr.net/Software/HEC/index.html Primary new Web pages:
Please see the section above on PMA Special Traces for a list of new traces pages. The NLANR.NET domain has been renewed for several years. The Web server (moat.nlanr.net) was replaced, and now has two 120GB hard disks in about 1/8th of its former chassis size. ~ Cichlid 3-D Visualization System Activities The effort to modernize and mature the current implementation the Cichlid 3-D Visualization System continued with work on the design and implementation of a multithreaded network protocol and development of a new class hierarchy. The idea of creating a Cichlid Web browser plug-in was investigated. This may be feasible as an ActiveX control, but probably not as a Netscape plugin; the idea has been shelved for the time being. Work continued in several areas, including establishing a better queueing system for network traffic and continued designing, building, and coding of the class hierarchy. Designed a ConnectionManager class for the server-side and worked on the DataSet class. The prototyping of a GUI for Cichlid was begun and a mock-up created. Currently working out how persistent storage of data will work so that users can replay their data streams repeatedly. A new dataset and graph code (not actual opengl rendering, just stubs for now) were written. Developed a new OpenGL rendering code for a pie-chart graph type. The images produced look very good and the code works very well. In addition to use in Cichlid, this code will generate images for publications, such as those created for the Citings/Proceedings Web page. http://moat.nlanr.net/Data-users/citings_proceedings.html Developed routines to do the drawing for vertex-edge graphs as well as routines to serialize and deserialize datasets. Began developing a more sophisticated interface for graph objects to publish their unique options (NURBS, render quality, interpolation) to the windowing interface in a graph-independent way. This consists of an interface infrastructure that can enumerate a list of graph characteristics and the available options (multiple choice) for each one, which then gets built into a menu. All of this happens without the window class knowing anything except a dynamically allocated data structure which holds character strings. Began creating a 'release' version of the Cichlid code. Several features still need to be completed (particularly support for labels and vertex-edge graphs). Began work on a collaborative project with HPWREN/ROADNet which involves developing a Cichlid server for the display of earthquake related data (animated). Wrote code that uses triangles to approximate an arbitrary terrain, such that a terrain map of San Diego County could be read in and displayed. Also wrote code that will animate a ripple-like effect over the surface of the terrain. It was tried on a couple of different terrains and the effect is quite stunning. Worked on the adaptation of the terrain map/seismic activity animation code into a Cichlid graph type. Now that the proof of concept has been completed, the biggest challenge is the generation of averaged normal vectors to make the shading of the surface look smooth and eliminate the natural stitching effect of the triangles. More detail on these activities can be found in the monthly and quarterly reports for this reporting year (April 2003 - March 2004), available at: http://moat.nlanr.net/Reports/
Student Research ExperienceChris Gross is the PMA student researcher. Chris had the opportunity of visiting New Zealand for the summer of 2003 and worked closely with PMA Project lead, Jörg Micheel, on a real-time passive measurement and analysis tool. This tool is intended to be deployed through the NLANR PMA infrastructure to, among other things, collect and process packet level statistics, information at the data-link, network, and transport layers, and eventually detect network "events" for the purposes of informing network administrators of problems on their link. The tool works with the DAG cards of Endace measurements to collect packets in real time. When Chris finished up his summer's work and returned to the U.S., Jörg wrote the following in his weekly report. "Overall, I am very pleased that we have managed to cover all the topics we had planned for Chris' visit to Hamilton [NZ]. He learned how to configure, access and control the DAG cards via the API, how to interpret the record formats, how to parse the packet data information, ARP, IP, UDP, TCP. He finished an application which delivers one second statistics on a variety of parameters that we had determined earlier as important. He managed to feed the stream of data into a small RRD database, and learned how to plot such data as a GIF file for Web access. These are all the steps needed from the lowest level all the way to the user interface for any serious real-time network analysis application. Well done, Chris." Chris Gross has concentrated on further development of the remote network sensor he began in New Zealand. He wrote a paper for PAM2004 about his work thus far. Writing the paper afforded him the opportunity to reflect on the project as a whole. He also learned much about his direction with sensor by explaining the reason behind and benefits of it. The paper really gave him an opportunity to fully see what he had done and the direction he needs to follow. He also worked directly with the Dag card designers which, upon his return to SDSC, has proved beneficial to other members of the PMA team. Lana Kennedy joined the NLANR/MNA staff in July as the Student Writer. Her skills with the formatting and organization of the progress reports grew quickly, which allowed her to complete more work to free up Maureen Curran, Technical Writer and Editor, for other responsibilities. She has learned to do more and more of the compilation necessary for the monthly and quarterly reports. She assisted Maureen in the creation of several Web pages, including the Citings: Data Users pages. She also formatted Mike and Ronn's "NLANR International Successes" article into two Web pages, and these were posted on the NLANR/MNA Web site. She had no prior experience with HTML or Web design, but adapted well to it and has become comfortable populating, editing, and even creating HTML documents. She also began a coding project, under the supervision of Tony McGregor, AMP Project Lead, in which she will add the Iperf bandwidth test to the new amplet reimplementation software. She has built her understanding of the UNIX operating system, network functionality, and C programming techniques. She had some experience with C++ and Java programming, but had no background in UNIX, and has progressively learned the fundamentals of the operating system. She was reallocated from the Iperf bandwidth project to a new programming assignment, a path display tool. The program's graphic output will display the path under investigation, and show portions of the paths it shares with nearby monitors to help locate problem paths. Under Tony's guidance, she began enthusiastically, splitting the project into two parts: a text output and a graphic interpretation of that output. She has learned much about C++ programming conventions and creating streamlined, efficient code, as well as how the AMP machines interact with one another and with the network. Lana also worked on the latest issue of the Network Analysis Times, an IPv6-themed issue. She compiled information from past progress reports for brief articles (which will also serve in Maureen's next design for the MNA home page highlights), designed the layout portion of the newsletter, and helped edit the finished version. Upon publication of this issue, Maureen wrote in her weekly report: "A huge thank you to Lana for her excellent, detailed work on the layout and other help." Working with Lee Dolan (Human Resources) and Maureen, Lana updated and formatted Hans-Werner's academic biography to meet the new standards specified by the Jacobs School of Engineering. This process was a good learning experience for her: she gained a better understanding of the differences between refereed and non-refereed conferences, and of the different types of academic works that are produced in the research community. Matthew Luckie continues to take the lead on two significant AMP projects, IPMP and AMP IPv6. His PhD work also continues to take shape. He developed and continued work on bandwidth estimation techniques using IPMP. The bandwidth estimation techniques are novel enhancements to current bandwidth estimation techniques, and are being tested for accuracy in the lab. He submitted a paper on this work (which was accepted by the ISMA BEst Workshop). As a result, he attended and presented at the invitation only Bandwidth Estimation Workshop, in December. The AMP IPv6 project was featured extensively in the latest edition of NA Times. He attended and presented at the CAIDA-WIDE workshop at ISI and the IETF in Minneapolis in November. He created and developed Scamper, the topology mapping tool (which is similar to Skitter). He also performed analysis on the data that Scamper has been generating. http://www.wand.net.nz/~mjl12/ipv6-scamper/ The Scamper project continues to evolve, with Path-MTU support for IPv4 and IPv6 as the latest feature. Matthew is looking at surveying the MTU status of links between sets of AMP monitors and providing this information to interested parties. Klaus Mochalski is a Ph.D. student from the University of Leipzig, Germany. His research focuses on real-time applications of passive measurements. During his three-month visit with NLANR/MNA last summer, he worked on two different approaches. First, he developed a new metric to characterize small-time scale network link utilization characteristics. He wrote a tool for measuring this metric in real-time and has used it to analyze some of the existing NLANR/MNA passive trace data sets. Second, he worked on an approach to identify bottlenecks along an Internet path by only watching the traffic on one end of the path using passive measurements. Therefore trace data is searched for signs of global TCP synchronization. This information complemented by routing information can yield a subpath containing the congested node. To achieve this, he began development of a tool to detect TCP flows and synchronization effects through real-time measurements. Apart from his research, Klaus contributed some time to the improvement of the PMA Web site (and has been very helpful in this regard). His technical expertise and knowledge of Dag cards and software has been a tremendous help as well. He may be returning this coming summer. More detail on these activities can be found in the monthly and quarterly reports for this reporting year (April 2003 - March 2004), available at: http://moat.nlanr.net/Reports/ Collaborations and activities supporting network researchWe continue our strong tradition of collaboration for both the AMP and PMA projects. The Network Analysis Infrastructure (NAI) serves as a platform with which we support our collaborators, and the high performance connection (HPC) and network research community, by making all of our data, analyses, tools, and techniques publicly available for use by network researchers, engineers, systems administrators, and students. As the size and scope of our various measurement and analysis projects has grown, we have expanded our outreach efforts to continue to meet the needs of the HPC community, as well as obtain feedback on future strategies and direction. We have numerous continuing collaborations, some of which we have maintained for years. The following list includes both long-term relationships, as well as newly developing ones. This year we continued, or began, work with the following researchers and organizations. PAM2003 ~ Our extensive efforts in hosting PAM2003: the Passive and Active Measurement Workshop, April 2003 resulted in a very successful and well received meeting. The impact to the research community was clear and it provided many excellent opportunities to develop and continue collaborations, as well as showcase our efforts. The meeting was well attended by the HPC network measurement and research community, including many students. One goal we had set for PAM2003 was to add some support for students and encourage their participation. These efforts were very successful - of 83 PAM attendees, 40 were students and 12 of the accepted papers had a student as the first author. A special handout on the AMP mesh and available AMP data, with suggested research topics was created for the student break-out session, which was well attended. Proceedings were published as print copies; author's papers are also available on the Web site. http://www.pam2003.org/ We are helping Jon Crowcroft and Ian Pratt, Cambridge with their planning for PAM2004. Reserved the pam2004.org domain name. Helped with preparations for the announcement of PAM2004, the Web page, and the necessary e-mail lists. Bruce Morgan, AARNet ~ AMP - They are developing an AMP mesh of their own. We have been discussing options with them for running their mesh as well as other aspects. They are awaiting the reimplemented amplet code release. Allied Telesyn, New Zealand ~ AMP - Met with the CEO (they do the development work for their layer 3 devices in New Zealand). Among other things, he has agreed to revisit the idea of them implementing IPMP. George Michaelson, APNIC ~ AMP - He is very enthusiastic about most things networking related. He is going to be running Scamper from some interesting locations within the APNIC region: Brisbane (AU), Japan, and Hong Kong. Nick Duffield, AT&T Research Florham Park ~ PMA - Requested data collected during the Slammer worm attack, January 25th. We were able to retrieve the traces of that day from the HPSS. The utility is excellent for this case: there are a number of other data collection projects which have not been in a position to help him in this case. The regular collection by PMA is clearly showing an advantage in postmortem forensics. He has also asked about IPv6 traffic traces. We do not have any such in our program as yet, although all traces contain enough information to gather the total amount of IPv6 packets and bytes (and relative to other data). Bill Cleveland, Purdue (formerly of Bell Labs) ~ PMA - Have been working together regarding mutual measurement activities and trace data analysis, particularly long traces. We provide the computing facilities at SDSC and make analysis graphs and data available to them (via the PMA Web site). One of the NAI computing engines has been dedicated to their use. http://pma.nlanr.net/Traces/long/bell1.html We worked with him regarding his measurements at Global Crossing. He is still keen on support for the instrumentation at a major provider on the East Coast. We recently discussed his success in being able to instrument a link at Global Crossing and new opportunities for joint work in collecting and publishing more trace data. Anukool Lakhina, Boston University ~ PMA - We sent three days of traceroute data (~450Mb) to him, he is a CS PhD student. MarkCrovell, Boston University ~ AMP - The new "day in the life" dataset was sent per request. CAIDA, k claffy, Margaret Murray, Brad Huffaker, Duane Wessels ~ AMP & PMA - We are supporting CAIDA's research work in a number of different ways, including working with them on the San Jose OC48mon pilot and collection of passive network data from a commercial backbone network at Above.net, SJC#2. (MFN - Metromedia Fiber Network is also a participant.) CAIDA is one of the most intense users of PMA trace data. Several research papers have been written and accepted at international conferences, including studies of longitudinal data. For the month of October 2002, download activities by CAIDA accounted for over 75% of the overall traffic at the PMA trace data server. We worked with them on our original development of Scamper. Duane Wessels was very helpful during this process. Scamper is an AMP IPv6 version of CAIDA's Skitter. It traceroutes in a parallel method to a list of IPv4 and IPv6 addresses. When combined with Univ. of Oregon's Route Views, and an extensive address list, it can be used to map the IPv6 Internet. Further development of Scamper is being funded by WIDE. Brad Huffaker has helped round out the existing address list with addresses from missing prefixes identified with routeviews6. He also ran Scamper and the address list from WIDE. As the main author of Otter, Brad helped fix a bug we encountered while working on the topology display, as we added options for displaying just the most common path each day and for omitting the tabular display. He was able to correct it and do a new release. This version prints even-sized nodes, which is good for our use of it. We have been working with Margaret Murray regarding the network infrastructure changes taking place at SDSC and the placement of regen taps at SDSC for passive monitoring. We meet regularly to discuss various measurement and analysis activities. Recently we had enlightening meeting with Margaret and the system administrators of CAIDA regarding SDSC machine room (and recent coming changes) in which a considerable amount of cooperative info was shared. Discussions focused around details of movement and consolidation of CAIDA, NLANR/MNA and HPWREN equipment: what equipment has been requested to be moved, what equipment is being removed, and what would be the better placement of such equipment. Jon Crowcroft and Ian Pratt, Cambridge ~ AMP & PMA - We worked with Jon Crowcroft and Ian Pratt of Cambridge regarding PAM2004. Have reserved the pam2004.org domain name. Helped with preparations for the announcement of PAM2004 and other aspects of the meeting. Director Yan Baoping and Kai Nan, Beijing, China ~ AMP - Soon they hope to complete the request to the government of China to get approval to host an AMP monitor. Another site in Beijing expressed interest in hosting an AMP machine, if possible. We plan to continue working with Director Yan. Bob Aiken, CISCO ~ AMP - Recently met with Bob; he expressed interest in IPMP and 10GigE measurements for AMP. Matt Davey, Corvil Networks ~ PMA - Cut a DDS4 tape with Auckland-VI data for him. Corvil Networks is a new startup in Ireland in the area of QoS-based IP networking. They wanted larger snapshots on tape and we determined which traces would fit best. Ana Preston, CUDI (Corporación Universitaria para el Desarrollo de Internet ) ~ The CUDI project in Mexico has an OC3 connection to CalREN at SDSC. The CUDI NOC is located at a university in Mexico City. An AMP monitor has now been deployed there. A possibility exists for further collaboration. Rasmus Hansen, Dante,Cambridge, UK ~ We were contacted by Rasmus about possible collaboration with them in measuring their network (GEANT2). We will meet with one of his colleagues at PAM. Christophe Diot, Intel, Cambridge UK (formerly of Sprint Advanced Technology Labs (ATL), Burlingame, CA) ~ PMA & AMP - Dialogs with him on his program to donate PC equipment for passive measurements; we continue to find a way to install more measurement points and support activities by other research groups. The IPMON research group is an example of passive measurement research in a commercial setting and we have exchanged ideas and explored opportunities for working together on a number of occasions. Sprint co-sponsored with NLANR the development of the prototype OC192MON at the University of Waikato. Sprint has also used PMA data for comparative analysis on a number of occasions, and published as research papers. Christophe has suggested that if we needed Sprint data we can use it at their labs, which is an offer to keep in mind. We held very good talks about ideas on future monitoring work. Xiaoming Zhou, Delft University of Technology ~ AMP - He has a project using hop counts from AMP data. ELTENET, Hungary ~ AMP - Met with Gabor, from Hungary, while he was visiting UCSD. They were interested in hosting an AMP monitor and possibly submitting a proposal to a European funding agency to deploy a local AMP mesh in Europe and make the data publicly available. An AMP monitor has now been deployed there. Stephen Donnelly, Koryn Grant, and Jesper Peterson, Endace ~ PMA - Worked with Stephen to make sure all the Dag tool options and statistics are proper for the new OC192 deployment and other issues related to the deployment of OC192MONs. Stephen Donnelly paid SDSC a visit in the Fall 2003. It was a pleasure working with him. Stephen will support integration works, and he also helped us with OC192 card testing and assembly into the DELL 2650 machines. Worked with Koryn to test Chris' Internet performance engine on a Gigabit Ethernet Dag card at Endace Knox Street. We have also been working with Koryn on a paper which looks at comparing a range of Gigabit network interface cards and their performance with each other. Jesper helped make changes to the anonymization tool to make it work for the GigE Leipzig traces.
Simone Triglia, Fiano ~ AMP - sent two months of data and the most recent "day in the life" dataset. Warren Matthews, Georgia Tech (formerly with SLAC) ~ AMP - Warren is a long time collaborator Recently, we set up am_master and sent him the am_slave code. Warren set up a SOAP Web services interface for some of the Internet2 data and has done something similar for AMP data, for which we are extremely appreciative. http://watt.nlanr.net/active/ampwebservices.html Tanja Zseby and Georg Carle, GMD FOKUS (German National Research Centre for Information Technology, Research Institute for Open Communication Systems in Berlin) ~ PMA - Tanja and Georg are active in the IETF IPFIX working group. We presented the PMA approach and discussed opportunities for monitor placements and data collection/publishing. GMD is an active user of the PMA trace data collection. The talk was attended by Dr. Juergen Rauschenbach, one of the key supervisors for network research at DFN, who is very supportive of strengthening the collaboration. Steve Corbato, Matt Zekauskas, and Rick Summerhill, Internet2 ~ AMP & PMA: We have worked with them on many collaborations. One current collaborative effort regards putting AMP monitors at each of the 11 Abilene backbone nodes as part of the Abilene Observatory project. http://abilene.internet2.edu/observatory/ The other is we have been working towards deploying an OC192/10GigE system into IPLS. We are ready to do the instrumentation; placement details are being worked out. We worked with Matt regarding the move of the ADV PMA monitor to Ann Arbor. In the Fall, we worked with him (and are grateful for his help) to prepare for, and install an OC192mon and an AMP monitor at the SC2003 NOC. We are working with him regarding the Ann Arbor Michigan PMA deployment at the I2 Gigapop. In the summer of 2002, we worked with Guy Almes, Caroline Carver, Steve Corbato, John Hicks, and Matt Zekauskas We are also working them on the E2E Performance Initiative, the Measurement Infrastructure Group, and other activities. Matt is looking into how to do authentication for active tests. Once he has that problem solved we might revisit the idea of doing some cross active mesh measurements between AMP and Surveyor. John Hicks, Jim Williams, Internet2 and TransPAC at Indiana University ~ PMA - Efforts to put AMP machines at the major GigaPops resulted in a request for an AMP machine from the Internet2 group for the Pacific Northwest GigaPop at Seattle. A machine was prepared, sent, and is online. Jim Williams discussed with us some of the near future plans for TransPAC. In a talk about his measurement activities, John Hicks mentioned the DAG cards and ENDACE and that Joerg has been instrumental in the development of the TransPAC passive measurements. They may want to have some additional active measurements on TransPAC. Eric Boyd, Guy Almes, Russ Hobby, and Warren Matthews, Internet2 Pipes Project ~ AMP - They are interested in using the AMP data as a fundamental building block of the PIPES E2E project. We have agreed that we would be happy to provide data to the this effort. Participated in the E2E TAG conference call. Recently discussed with Eric Boyd and Warren Matthews about getting Warren's Web services code onto AMP. We are also trying to organize a meeting to discuss whether there is anything that we can to between OWAMP and AMP. Bill Manning, HPCC Division, ISI ~ AMP - Provided data to obtain a route-view to compare the address list against what is actually routable, and to strip out AS numbers. We were then able to run Scamper over the address list and produce a logical map of the IPv6 Internet. Manhee Lee, et al., KISTI/KREONET ~ PMA & AMP - We received a request for a PMA monitor from the KISTI site (Korean Institute of Science and Technology) in Taejon, Korea. The monitor was prepared (using the newly arrived latest version of the Endace software) and shipped to the site. KISTI is the HPIIS hub for Kreonet, the Korean R&D network, and a StarTAP/Abilene partner. MNA has worked with the KISTI folks on active measurements since the HPIIS workshop in San Diego (August 2001). We are providing information and some support regarding bringing KREONET's new AMP machines into the AMP mesh. An article on their AMP activities appears in the November 2002 issue of the NATimes (3.2). This will provide us with the opportunity to leverage the existing NLANR AMPs in APAN countries to help create a larger APAN active measurement infrastructure in collaboration with KREONET/KISTI. Joe Abley, Internet Software Consortium (ISC), Auckland Peering Exchange ~ AMP and PMA: Have been in touch with them regarding support of measurements for the new F.root mirrors and DNS server configs that are being built around the world, and how we can help. The AMP machine at the Auckland Peering Exchange is online. Surasak Sanguanpong, Korea ~ AMP - Surasak has been speaking with us about setting up an AMP mesh using their own hardware but our software. Brian Tierny, Lawrence Berkeley National Laboratory ~ AMP - Spoke with him about the GGF measurement standards. He was happy to hear this standard will be used for AMP data. Larry Blunk, Merit Network ~ AMP - Sent us copies of his IPv6 slides, which were excerpted into a short article for the Network Analysis Times. Chuck Blake, Dina Katabi, and Sachin Katti, Massachusetts Institute of Technology (MIT) ~
PMA - Our data was used (en masse) in "Cross Traffic: Noise or Data," their paper was presented at Internet Statistics and Metrics Analysis (ISMA) Bandwidth Estimation Workshop, San Diego, California, Dec. 2003. "To assess the distribution of cross-traffic in the Internet, we studied 300 million packets in 162 NLANR traces that spanned several months of traffic on one OC12 and eleven OC3 links. We identified 28,000 significant flows." http://www.sds.lcs.mit.edu/~dina/dkpub.html John Storm, Morgan Stanley ~ AMP - He moved from an HPC site and is trying to interest people at Morgan Stanley in AMP. David Hanley,National Center for Data Mining (NCDM),University of Illinois at Chicago ~ Interested in the new TeraGrid 10GigE/OC192 traces. William Maton, National Research Council, Canada ~ AMP - Working with William F. Maton - who has been collecting Scamper runs for about 4 or 5 months now from one source to a constant target list - has generated data that will enable us to compare IPv6 routes that have changed over time. He has started using a variation of the script to generate IPv6 Scamper traces automatically. We started doing some analysis on the weekly Scamper traces that William Maton is generating for us. http://www.wand.net.nz/~mjl12/ipv6-ryouko/ Using the latest Scamper run that he produced, we found 178 alternate paths. A few were probed with the 'ring' tool to check that it works. Greg Cole, NaukaNet ~ AMP & PMA - We attended the RELARN meeting in Russia with Greg and continue to discuss his interest in AMP and PMA, his needs for monitoring, and how we are going to approach them. Set up an amp-naukanetnwu to test to a set of machines in Russia, to which Greg Cole arranged for us to test. Greg hopes this will be a stepping stone to a Russian AMP mesh (as do we). http://www.gloriad.org/ Working with him on possible measurement infrastructures in Russia and China. We are also working with him regarding the Gloriad project.
Pekka Savola, Netcore ~ AMP - His internet-draft prompted work on finding possible examples of /127 routing prefixes--it is now RFC 3627. He provided a few pointers on how we could find routing prefix lengths. John Towns, NLANR/DAST ~ Met with him to discuss the nextinet project and NLANR collaborations. Video teleconference discussing potential collaborative projects for NLANR DAST and MNA. NLANR managers discussed using AMP data for DASTs advisor system. The thinking we have been doing about implanting the GGF NMWG XML schema for measurement data would assist this project too. There are still some hurdles for us to get over, but it is clear this is an important direction. Yuri Demchenko, NLnet Labs ~ AMP - Sent him a copy of our slides from the RELARN meeting, as requested by Greg Cole of NaukaNet. David Kessens, Nokia IPv6 ~ AMP - He is very involved with 6bone, and we've worked with him to discuss methods to construct an IPv6 address list. Cedric Westphal , Nokia IPRG ~
(formerly of Ipsilon) Hank Nussbacher, Israel InterUniversity Computation Center, Tel Aviv ~ AMP - requested on through info on the I2 measurement mail list. Bill Owens, NYSERNET ~
AMP IPv6 - He is one of the people who is working closely with us on (and encouraged) the AMP IPv6 measurement and analysis mesh from its inception. Bill has contributed two articles for the Network Analysis Times (NATimes). Bill Owens has offered to help us out with a 9k MTU clean path by installing an OC3 ATM card into a Linux machine and plugging that directly into a router. Paul Schopis, OARnet GigaPop ~ AMP - Met with him; he is supportive of hosting an AMP at the OARnet GigaPop. Casey O'Leary, Pacific Northwest National Laboratory (PNNL) ~ PMA - Casey would like to offer his support in NLANR's OC192MON project for SC2003. We previously worked with him at SC2002 regarding an OC48MON. Peter Arzberger, Teri Simas, PRAGMA ~ AMP & PMA - We continue to work with the Pacific Rim PRAGMA initiative, the focus of which is to create and strengthen ties with Pacific Rim nations. PRAGMA will be leveraging UCSD, SDSC, and aspects of NPACI. Our ongoing international collaborations put us in a good position to assist this effort. We hold regular meetings with Peter Arzberger about PRAGMA upcoming events and activities. Brought him up to date on Greg Cole and the NaukaNet project. Attended his PRAGMA meeting. Ronn gave a presentation with him to Dr. Yang from Korea. Also worked with him on the CUDI meeting in Mexico in early October, where he showed a couple of NLANR/MNA slides for us in his presentation. He also pointed out that the CUDI NOC in Mexico City is hosting an NLANR/MNA AMP machine. Met with him about PRAGMA funding and future plans. The brochure that PRAGMA has included a number of references to our work in a brochure on their project. We have provided them with text and images. The next PRAGMA meeting will be in China in May 2004. Peter and Greg are to meet with Director Yan from the CNNIC site in Beijing - which expressed interest in hosting an NLANR/MNA AMP monitor and is participating in the Gloriad project.
Alexandre Grojsgold, RNP(Rede Nacional de Pesquisa), Brazil ~ From RNP in Brazil talked about hosting an AMP and possibly a PMA machine. He has had requests for packet header trace data from researchers. An PMA machine in Brazil can make that data available to those researchers as well as others. He is also interested in having some graduate students work with the existing AMP code to create a local AMP infrastructure in Brazil and then making that data publicly available (as is all of our data). The idea is that this group of graduate students can provide feedback on any modifications and/or additions to the existing AMP software. An AMP Monitor is now deployed there. Lorenzo Colitti and Michael Swoboda, RIPE NCC ~ AMP - Lorenzo implemented path MTU discovery through the RIPE NCC TTM boxes and uses it to find IPv6 tunnels (which is the same idea behind what we are doing). He pointed us at his code and a great technical report they have written on it. Michael sent the output from running Scamper on what seems to be all the RIPE TTM boxes that are IPv6'd. He sent the output from 20 Scamper runs. It may be possible that we now have the bulk of the European IPv6 Internet mapped thanks to Michael. Henri Casanova, SDSC ~ Henri requested that we send him slides about our activities with the goal of attracting UCSD CS grad students (he gives this SDSC recruitment talk twice a year). Vijay Samalan,Networking Director, SDSC ~ AMP & PMA - Vijay is the new Networking Director at SDSC. He asked to see how NLANR/MNA might be able to help with performance monitoring tools for Grid computing and for us to sit in on a few meetings with Grid people. Hopefully with the existing infrastructure, we can provide some needed data. Recently we discussed with Vijay Samalan about collaborating with the OptiPuter project. Jay Dombrowski, Kevin Walsh, SDSC Network Operations group ~ AMP & PMA - Worked with him to share utilization of the Spirent Adtech equipment. We used it to provide signal generation for the OC192 monitor. Great enthusiasm was shown in assisting us with this project. He also helped out at SC2003 with the AMP monitor. Also worked together regarding the SC2003 Bandwidth Challenge. Worked with them regarding the network infrastructure changes taking place at SDSC and the placement of regen taps at SDSC for passive monitoring. Also, Kevin asked to be put in contact with Jim Ferguson from DAST so they could talk about NPACI E2E measurements. Worked with Jay, Kevin, and Tom Hutton re the connections of the two passive monitors for the Abilene and CalREN connections at SDSC. Also worked closely with them to implement an IPv6 connection for the AMP machine in the SDSC machine room. Kevin has assisted us in the configuration of the Dag card test bench. Phil Papadopoulos, SDSC ~ Discussed possible future OptiPuter measurements. Currently a testbed has been implemented. Les Cotrell,Warren Matthews, Connie Logg, Stanford Linear Accelerator Labs(SLAC) ~ AMP - Spoke with Les Cotrell about 10 GigE cards and perhaps working with us on a 10 GigE AMP. He has been doing evaluation of 10GigE interface cards. Sent Warren the am_slave code and wrote up some installation instructions so he can get the data from the SLAC monitor in real time. We also discussed event detection and an IPv6 monitor. In particular they are interested in the event detection work done a couple of years ago. They plan to use the algorithm as part of their bandwidth estimation work. Randy Presuhn, Simple Network Management Protocol version 2 (SNMPv2), RFC Working Group ~ AMP - He suggested changes to the IPMP Internet-Draft. Randy is editor of the SNMPv2 RFC. Heonkyu Park, South Korea ~ AMP - Helped him get IPMP going on his FreeBSD 4.3 machine. Henk Uijterwaal (RIPE-NCC), Wim Biemolt, Surfnet.nl, Ronald van der Pol, NLnet.nl, and Tim Chown, University of Southampton ~ AMP - Sent them Scamper to run, which will provide views of the IPv6 Internet from machines in the European Union. Awaiting output from them. Taiwan ~ Che-nan Yang (Director, Taiwan) and his deputy director, Dr Tsai, Expressed interest in the AMP project. Taiwan has some ambitious network upgrade plans and would like to include network measurement as an integral part of that effort. They expressed interest in hosting two AMP machines - one at their Northern NOC and one at their Southern NOC. This group from Taiwan may be the third country to follow the new collaboration model by hosting an AMP monitor and deploying their own local AMP mesh. They have already started experimenting with the current AMP software. They are making progress with their local AMP mesh, and we are assisting them as needed. They have one amplet and a data collector testing to our international mesh. We've been working closely with them to get their monitor integrated into our AMP mesh. Sent the latest international list, and created a Web page where that can be accessed (using a username and password). Fey Sheu, Taiwan ~ They will develop their own AMP infrastructure at the GigaPOPs and possibly further onto campuses. A delegation from the National Center for High Performance Computing of Taiwan visited SDSC in December. Taiwan will soon be upgrading their network infrastructure to 40 Gigabits. Interest was expressed in hosting one to two AMP monitors and in developing some local measurement capabilities. We met with them and gave them a tour of SDSC, including the machine room where our local machines are located. Thailand ~ Dr. Wanchai Rivepiboon, Chulalongkorn University Asked to continue the dialog on measurements. Joerg has arranged for an OC3MON to be placed on the Internet2 connection to Thailand. Dr. Rivepiboon is interested in hosting an active monitor and asked Ronn to give a talk to a local measurement workshop. The UniNet group from Thailand invited us to their December meeting, which may fit well with another project in Thailand that wants to leverage NLANR/MNA measurement work in that country. An AMP Monitor is now deployed with UniNet. Jim Williams, TranPac ~ He is interested in having NLANR/MNA continue the measurement collaboration in their next proposal. Weidong Chi, University of California, Berkeley ~ AMP - He has collected all our online data (via the Web get interface). He wanted historical data from around the code red time, so we wrote some code to extract that from the HPSS. He also requested to be enabled for the on-demand throughput tests. We continued sending more data during November, including the 5 months' data he requested. Yan Chen , University of California, Berkeley ~ AMP - Asked for two months of AMP data this period for use in his group's investigation into adapting Internet application |
