NLANR Measurement and Network Analysis Group (NLANR/MNA)
Passive Measurement and Analysis (PMA)Indianapolis (IPLS) router clamp installation (OC48) ~ Abilene - PMA project lead, Jörg Micheel traveled to Indianapolis for the installation of the OC48 PMA machines at the Qwest POP on Kentucky Avenue. He worked with John Hicks (of TransPAC at Indiana University), Caroline Carver, and Grover Browning (both of the Abilene Team at Indiana) to bring the splitters into the links to Cleveland (CLEV), Kansas City (KSCY), and Atlanta (ATLA), (see http://abilene.internet2.edu/maps-lists/). The OC48c link to Chicago was considered too fragile to be touched. They completed the installation and connectivity for the two OC48mons looking at IPLS-CLEV and IPLS-KSCY. There were technical issues with the link to ATLA which made it unsuitable for any serious monitoring. Via emails with Steve Corbato (Internet2), it was determined that a further instrumentation of the IPLS router would not be feasible at this time, and that this level of installation would be all that could be achieved given the various constraints. Trace collection with the pair of OC48 monitors at Indianapolis (IPLS) was accomplished after the MOU with Internet2 was finalized. Work began on providing stratified sampling from these Indianapolis OC48c machines. The plan was to implement a systematic sampling scheme similar to the other PMA monitors, perhaps slightly longer snapshots, which we ran until the links at IPLS were dismantled. A two-hour snapshot from 9am to 11am CDT was collected. The first of the Abilene OC48c data, Abilene-I, was published after the trace files were post-processed (anonymized, etc.) and all of the standard trace graphs (packet, bandwidth, flows, volumes, etc.) and the .html pages were generated. Abilene-II, which is a collection of five minute stratified random samples, eight times a day, was published a bit later. It is almost identical to the rest of the PMA monitors, except that we use different tools and also publish five minute samples, which amounts to a good 1GB of data each time. Consequently we ran out of disk space, therefore not all data is available directly from the Web. A listing of the additional data available is on the Web page; this data can be requested by email.
The same day of the two-hour snapshot from the OC48 monitors at Indianapolis (IPLS), another snapshot was cut, independently, for CAIDA folks in San Jose (OC48mon pilot), eight hours. All together, the traces collected during this one day comprised a total of 500GB of OC48c packet headers (our first at this speed). What a day for passive Internet measurements! The eight hours from San Jose is probably the largest IP header trace file in the history of passive measurements, with a trace volume of 336GB uncompressed. That's 5.6 billion IP packets, contiguous and CDMA synchronized. The sheer volume raises many questions about the usefulness. OC192 Measurement and Analysis - In November, we successfully acquired and implemented a platform for an OC192 monitor supporting an Endace Dag6 card. We now have an OC192 monitor platform in place, connected to the DTF OC192 link to Chicago (located at SDSC). Unfortunately, there were no systems connected to this link, therefore no traces could be taken nor tests run. OC192MON activities continued making excellent progress in early 2003. Jörg went to Paris in January, where he worked with Christophe Diot of Sprint ATL on the field installation of the first OC192MON, tapping into the link between Paris and London. The work was successful, we managed to collect a few hours of trace files, with GPS synchronization. Link loads were between 420,000 and about 510,000 packets per second, and around 2.5 GBits/sec, equivalent of a full OC48c. This is an average 600 Bytes/packet. Work is continuing with this pilot in Paris; we hope to get SDSC involved very soon in order to have a better testbed with access to test gear. Development of new metrics - New PMA student, Chris Gross, spent time learning about the PMA trace files, specifically the tsh format. He and Jörg discussed plans for how to proceed with some interesting new analysis work which will eventually involve permanent monitoring and a low-volume data stream to be displayed as RRD graphs on the pma data server. The plan is to also develop real-time analysis for PMA. Chris worked on developing a set of passive measurement metrics. Additional Passive Measurement activities Stats analysis of pma.nlanr.net (Web log) - The trace data download log files have been collected and integrated with existing files to create an historical report. A tool to filter httpd and ftpd logs into a common format and merge into a single sorted file was scripted. We are using awstats to generate reports of trace download data; work continues on the interface. Work on displaying the http/ftp download volume of PMA data has revealed that overall, the old statement holds true that the amount of download activity by users matches the amount of data collected, in the 3-5GB per day range. This is good news, as it reflects that, on average, all our data is useful, is being looked at, and studied. Work continued on the PMA sampling research. A summary Web page that displays data for sampling rates of 4, 8 and 16 was created. Phase space analysis on traces: code was written to create a 3-D representation of the analysis of a phase space on the interarrival times of packets in a TSH file (PMA trace data). Having successfully done that, experimented with creating the phase space based on the packet length and adding additional information to the visualization using color to represent size classes. Obtained some interesting results, the meaning of which is not yet known. ~ Status of PMA sites and new deployments ~Significant progress was made on getting the outstanding PMA monitors back online and collecting data. We worked with Jay Dombrowsky and Kevin Walsh (SDSC Network Operations Group) with regard to connecting passive monitors to the Abilene and CalREN connections at SDSC. The machine for the Abilene connection located at SDSC (nai-p-sda) was completed, placed back on the connection and is collecting traces. Discussions continued regarding the connection to the CalREN network. When the latest software CD from Endace arrived and installation was successful on the SDSC Abilene machine, the Tel Aviv (TAU) and Front Range GigaPop (FRGP) monitors followed. We worked with the Tel Aviv site people planning the replacement of their PMA monitor with one that can monitor their new POS connection. FRGP also required a PMA monitor replacement due to an upgraded connection at the site. The OC48 monitor for the Mid-Atlantic Crossroads GigaPop was shipped. Bud Hale is working with Dan Magorian at the Mid-Atlantic Exchange GigaPop (MAX) who has the new OC48 monitor in place at the site in downtown DC. (They were experiencing some ethernet problems in the facility, but are expected to have it back online beginning of the next period.) We are planning to create a GigE PMA monitor at the Google GigaPop in the northwest. Received a GigE PMA monitor interface card from Endace. A very interesting possibility arose when a PMA monitor was requested for placement at the connection to APAN at KISTI in Korea. We are preparing a machine and awaiting configuration data. A new passive measurement monitor is also being prepared for installation at StarLight. We communicated with Brent Sweeny at Indianapolis regarding the OC12MON and the three OC48MONs there, seeking to place a new set of monitors at the new Qwest POP. Another development on the PMA scene is that NCAR (National Center for Atmospheric Research) is requesting a GigE monitor. There are a few questions that need to be answered but that machine should be taking shape early in the next period. An inventory of the NLANR PMA equipment and devices currently on hand was performed, so we know what is available for use in future PMA deployments. More detail on these activities can be found in the quarterly reports for this reporting period, available at: Active Measurement Project (AMP)More nations are realizing the importance of high performance (HPC) research and education networks. The number of national HPC networks has grown significantly over the last few years. Many national HPC networks are interconnected with those of other nations, including the U.S. based HPC networks. Many of the challenges that face intra-national HPC networks are more severe in international networks. As HPC networks spread around the globe, the need for advanced network measurement increases both in the number of applications and, more importantly, the nature of the measurements required. The AMP mesh now has approximately 130 machines in it. AMP made progress towards extending into new measurement domains this reporting period particularly with regard to new international deployments (Tokyo, Japan (APAN link), ELTENET in Hungary, HEANet in Ireland, TCTnet in Finland, and UNINet in Thailand) and the development of a new collaboration model where organizations develop and run AMP meshes of their own with some limited support from us. Two such meshes are the AARNet (Australia) and Kreonet/KISTI (Korea) meshes. The creation of meshes within the AMP infrastruture and the development of a new testing archtecture were significant activities in support of these efforts to extend AMP deeper into the HPC community. New testing architecture - We have reached the point where it is difficult to manage the different tests that we perform on the AMP mesh with the ad hoc approach that we currently use. Therefore, we are developing a new testing architecture for AMP which will allow us to add new tests and control which monitor does which tests from the database. The new architecture will add support for a more flexible range of tests, meshes, and test schedules. This is important because we want to extend AMP into a range of new tests, including the IPv6, and regular bandwidth tests, etc. Not all sites will want, or be able, to support all tests. Therefore we need a robust way to manage which site does what. The challenge is to create something relatively simple that accommodates what we need, with new types of tests being added. One interesting aspect is the need for some tests that are done only if others cannot be run (e.g., use ICMP if you can't use IPMP). A proposal (guidelines) for how we can extend the AMP testing architecture to accommodate these needs has been developed, and is being used as a guide to implementation. The Web pages and database to create new tests were completed. An important detail not included in the original design document is how to handle defaults. It is important to be able to enable, disable, and set to the default the ability to do a test. This is mostly for implementation reasons, but may also be useful, for example if we upgrade a test and this requires more abilities from the monitor. IPMP - The IPMP Internet Draft was completed and modified to reflect the current use of the protocol. It was submitted to IETF and is posted on the IETF Internet-drafts directory. http://www.ietf.org/internet-drafts/draft-mcgregor-ipmp-02.txt Development and modification work on IPMP included substantial changes to the format of the echo packet, the addition of an IPv6 path record structure, flow-on changes to the information packet, and the addition of a real time reference point structure. Changes and improvements were made to IPMP including changing the path record structure to contain the TTL value in addition to the receiving IP address and timestamp. The information packets were also updated to supply an error bounds with each real time reference point. This is important for the work we wish to do with NTP on the AMPlets. The quality of the various NTP servers (that sites have advised) was investigated. Additional work was done regarding IPMP protocol changes to do bandwidth estimation on half duplex networks (e.g., wireless networks). The crux of this work is to give the protocol the ability to measure in just one direction and return measurement results out of band. Dispersion Analysis of an IPMP Packet Train is shown in: http://voodoo.cs.waikato.ac.nz/~mjl12/. One-way delay measurements using NTP were conducted with a few sites from sorcerer, the machine located at Waikato University (New Zealand). AARNet's (Australia) path to Waikato (across the Tasman Sea) produces very interesting graphs ( http://moat.nlanr.net/~mjl/graphs/aarn/) where we can see that the variability in RTT is due almost entirely to the reverse path back from Australia into New Zealand. We plan to send a CDMA time receiver to one or two AMP sites to support the one-way delay using NTP activities; the first went to NYSERNet (New York). We do not have plans to equip many AMP machines with time receivers, as they are expensive. A couple will give us a starting point and a good reference with which to compare. One of the major accomplishments for IPMP was the taking and analyzing of measurements with IPMP on the CRCnet (Connecting Remote Communities Network, http://crc.net.nz/) wireless network, with an aim of estimating the capacity of the bottleneck link. In order to achieve this, a Linux implementation of IPMP was created so that it could be deployed on CRCnet. The dispersion graph ( http://voodoo.cs.waikato.ac.nz/~mjl12/dispersion.png) is very interesting because we can see contention at the end-point of the network. The big gap between MCG, that begins to close up at PIR on the reverse path shows two artifacts of 802.11b: 1) each packet requires an acknowledgement to be sent back, and 2) L2 is half duplex (one radio can speak at a time), which means the packet-pair becomes separated two hops back from the target. IPv6 - Discussions and planning for the development of an IPv6 mesh within AMP were held in July and August 2002. In part, we pursued this addition to AMP as a response to the needs of the HPC community. With the creation of our IPv6 mesh, we perform measurements which enable systems administrators and engineers to monitor infrastructure improvements to the IPv6 component of Internet2, as they occur. We can also compare basic measurements of performance to that seen with IPv4. Bill Owens (NYSERNET) and Joe St. Sauver (U. Oregon) have been significant collaborators in this activity since the beginning. Thanks to Bill, the IPv6 working group for Internet2 is aware of our work with IPv6. A number of AMP related projects were performed in order to make IPv6 possible, including an IPv6 KAME implementation for IPMP, making ipmp_ping follow the KAME guidelines for writing a dual stacked application, and traceroute6 tasks. Each AMP site will have the traceroute request software working and IPv6 turned on by default (for autoconf purposes). An email invitation was sent to all AMP site administrators, asking if they were interested in participating in the IPv6 trials (by assigning their AMPlet an IPv6 address, the monitors end up with two addresses). The email also invited sites interested in the NTP based one-way delay work to contact us. After NYSERNet was configured for IPv6, we sent a CDMA time receiver for placement there which will allow accurate time delay tagging at that site, as well as providing a reference point for the one-way delay measurements. We worked with SDSC Net Ops to acquire an IPv6 IP address for the amp-sdsc machine. Although IPv6 exists at SDSC, it was not easily connectable; however, the connection was successfully completed. Also a CDMA time receiver was connected to the SDSC AMPlet to provide for delay measurements. We do not have plans to equip many AMP machines with time receivers, as they are expensive. A couple will give us a starting point and a good reference with which to compare. In September, we had a couple of machine IPv4 address changes as a consequence of people wanting to move their monitors to put them on an IPv6 network. By mid October, the mesh was up and running with eight sites having an IPv6 address. Also, at that time we started collecting IPv6 measurements; the scripts to collect the data and generate those pages were written and the traceroute grapher was ported (to understand the traces being collected). http://watt.nlanr.net/active/cgi-bin/v6_portal.cgi In early December, we began analysis of IPv6 data over the six week period during which we had been collecting data. Graphs were generated which showed how the distribution of IPv6 RTTs changed over time. By looking for sudden shifts, several interesting cases were found. Code was written for the traceroute grapher in order to position it for use with AMP IPv6:IPv4 comparisons. Several modifications were made to the code initially written by James Spooner. A planned route change that occurred in the direction of the University of Oregon can be seen at: http://watt.nlanr.net/active/cgi-bin/v6_linkcomparison.cgi?from=amp-uoregon&to=amp-nysernet&date=102.12.17. AMP IPv6 / IPv4 comparisons - Originally, static pages were generated every 15 minutes displaying these comparisons; then later in the period, there was a change so that comparisons could be generated upon request. Example IPv6 / IPv4 comparisons:
An IPv6 address list to trace the forward path to all the IPv6 prefixes has been started. The IPv6-world trace has 860 addresses (1141 prefixes). Matched each address, where possible, with a prefix and have 645 addresses matched to a prefix. Unfortunately, 311 prefixes have one or more addresses, out of a total of 1141 (or about 27%). Discussions with Joe St. Sauver and Bill Owens have included the possibility of developing plans to try and map the MTUs seen between select AMP paths. To do so, we will need to install GigE cards that have an MTU of approximately 9180 bytes, and get sites to be jumbo clean all the way to Abilene. They can do this in a relatively short time, but other sites might be a problem - if they are even interested. The choice of GigE cards is important because apparently many cards currently in production fall (way) short of the needed MTU of approximately 9180 bytes. The AARNet machines for example, have been shipped with the 3Com 996B-T cards, which are likely 4470 byte MTU cards. There are currently 11 sites in the IPv6 mesh. We have monitors in all corners of the US, one in Japan, and one in Australia. This coverage is sufficient to see important backbone upgrades to Internet2, according to users of the data. However, we would like to have IPv6 monitors in Europe. We have no plans to specifically recruit additional sites to join the mesh. We anticipate that sites will join as they find a need for the information. Other groups have recently begun looking at the IPv6 Internet, but as far as we know, we are the first to create an IPv6 mesh, produce this type of information, and make the data publicly available. http://watt.nlanr.net/active/cgi-bin/v6_sitesforday.cgi?date=103.3.31 Scamper - is an 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. (We are collaborating with CAIDA on the development of this tool.) The general architecture of this traceroute utility is a tool that can conduct multiple traceroutes simultaneously so that if it gets stuck on one hop (a hop that does not generate ICMP destination unreachable messages), it can traceroute other paths until that hop times out. The time between each probe can be configured so the number of bytes per second this tool generates is configurable. OS upgrade - A significant project on the AMP system was the updating of the operating system (OS) on each of the AMP site monitors (AMPlets) from FreeBSD3.0 to FreeBSD4.6. After the upgrade the AMPlets received OpenSSH 3.2, which was a recent addition to the FreeBSD -stable branch. There were several reasons that the OS upgrade became a high priority. Site administrators were worried by the old system and it was becoming harder to get security fixes and support for the old OS. And, importantly, there are several things we want to do, but cannot, with 3.0, such as IPv6, one-way delay with IPMP and NTP, GigE, as well as other things. The upgrading of the AMPlets was an almost site by site exercise with each process being run interactively. Despite this, it went quite smoothly in general. This huge task included developing, then extensively trialing the process, followed by an extended period of actually upgrading the monitors, and troubleshooting the process. We did not lose a single AMP monitor during the upgrade. However, several machines (approx. 12) did require individual attention. These stragglers were brought up as their individual issues were discovered and corrected. We then updated the system manager to 4.6 (not a trivial process). A list of the seen transfer speeds for downloading a 145MB /usr tgz to the AMPlets from moat can be viewed at: http://moat.nlanr.net/~mjl/upgradespeeds.html. By the end of October, we had achieved almost 100% completion of the OS upgrade for the AMPlets. An infrastructure-wide system update of all FreeBSD4.6 AMP remote sites was performed after multiple changes were made to the master system disk and successful testing. Also, most of the issues re the master (FreeBSD 4.6) system disk had been resolved, which will allow the preparation of many new machines for shipment to remote sites that have previously requested them. Many discussions were held regarding upgrading AMP and VOLT (the two data collection servers for AMP) to FreeBSD 4.6. At the same time, an update of the Apache server software is planned. Rather than just upgrade the machines, we plan to create a new (third) server and migrate everything to it. When it is operating (with a fresh 4.6 install), the existing servers can be upgraded. This upgrade strategy was chosen in part because AMP and VOLT are no longer well configured (e.g., their root partition, at 2Gb, is too small). And, as a first step to the repackaging/reimplementation of AMP, a list of everything that is installed on the machine is being maintained. In keeping with the trend to physics parameters and inclusion of the remaining term in Ohm's Law, this temporary server will have the host name of "ohm.nlanr.net." ~ Other significant AMP activities ~Major infrastructure work during the period included the mesh creation and writing code to create the ability to delete a mesh from the database. In addition to improving the current mesh management, these activities provide tools that will be useful as the number of meshes increases with campus deployment and the development of other meshes which will extend AMP. In addition, an international mesh with all the non-US sites in it as well as sdsc, psc, and bu was created. Over time, we would like to migrate the international sites out of the main HPC mesh. Work restarted on event detection, primarily on establishing the status of the project and documenting more of the code. Investigation of why it is apparently not detecting an event that it should, was started. Currently the algorithm is complicated, but so far nothing simple works. A visualization project for AMP to dynamically render sites and serve as a basis for rendering real-time performance of AMP sites was begun and significant progress made. A new "day in the life" dataset for AMP was created, distributed to those who had made recent requests, and posted. The AMP system disk was modified per a decision (in December) to make it such that it has a network card driver installed for the Intel NIC, RealTek NIC, and the 3COM 3C996-T gigabit network card. This system disk version will be shipped with all AMP monitors in the future, including the monitors for Australia. We also resolved an airflow issue with the 1 RU AMP chassis in consultation with the vendor. The decision is to create a vent in the cover at the vendor location by opening a hold in the cover with a 2 inch metal punch and cover it with screening. We have decided that we should change from using mSQL (which is proprietary) to postgresql for the AMP support; there is no timetable for this change as of yet. It has been decided that we will stop collecting temperature and voltage sensor data as it was only available on a few machines and does not seem to have actually been useful in detecting system faults. ~ AMP remote sites: new deployments and status of sites ~New AMP monitors were prepared and shipped this period to the following sites: Tokyo, Japan (APAN link) U. of Idaho, Stanford Linear Accelerator (IPv6 monitor), ELTENET in Hungary, HEANet in Ireland, TCTnet in Finland, UNINet in Thailand, U. of Wisconsin Network, and Great Plains Network GigaPop. Six additional monitors were sent to Australia for the AARNet mesh. With minimal exceptions AMP sites are collecting and transferring data very solidly. AMP servers are running well. Archiving of the VOLT server disks was successfully performed (due to previous problems, the process was carefully monitored). Even though we have not migrated to the recommended HPSS interface, it appears our archiving issues are mostly resolved. Late in the period, the new AMP server, ohm, was completed and installed in the machine room at SDSC. More detail on these activities can be found in the quarterly reports for this reporting period, available at: Additional Performance Measurement and Analysis ActivitiesWith the latest installations to HPWREN on San Clemente Island (about 72 miles off coast) and the King Stormwater Bridge near the Salton Sea, HPWREN now literally stretches from the desert, through the mountains, via the beach, and ways into the ocean. ROADNet's link to the R/V Revelle goes out into the high seas. An updated topographic image of HPWREN is available at http://hpwren.ucsd.edu/topo.html. HPWREN measurement and analysis activities this period included radio signal event detection, network event reports and rebuilding the PAM machines to do different measurements. (The HPWREN measurement machines are configured for both passive and active measurements, therefore they are called PAM machines.) Several of the measurement and analysis Web pages were improved and updated, including the UPS pages. The Tsunami event detection script was completed and documentation added. There is still work left to do for automatic user notification of events. More progress was made on the network utilization script (for the network reports page and email). The anemometer data graphing tool originally created in the last quarter, was improved. (Note that this tool has since become unusable and as yet, cannot be revived.) More detail on these activities can be found in the quarterly reports for this reporting period, available at: Documentation, Networked Data, and ToolsExtensive groundwork and study has been performed re the current state of the Cichlid 3-D visualization system, with a view to both reimplementing it, as well as developing new functionality. An aspect of this was a project to research the current state of the art in the rendering of multidimensional datasets. The new Cichlid/ visualizations student, Ben Reesman, performed this research and presented his findings internally. The development of Cichlid has been dormant since the creator and original developer, Jeff Brown (then an undergraduate researcher with us), left after graduation to begin the UCSD Computer Science Ph.D. program. Ben has opened a dialogue with him, discussing with him the foundations and reasoning behind aspects of Cichlid. Work has begun on updating the current implementation, on creating a cross platform GUI, and on reimplementing it in C++. As mentioned previously in this report, the Abilene-I and Abilene-II OC48c data were published. Abilene-I: http://pma.nlanr.net/Traces/long/ipls1.html Abilene-II: http://pma.nlanr.net/Traces/long/ipls2.html A world map highlighting NLANR/MNA international collaborations for both AMP and PMA was created. This new world map impressively displays the depth of our collaborative efforts; and it will be used for (slide) presentations. In conjunction with the major infrastructure work on AMP regarding mesh creation, a Web page was created to list the AMP sites that are a members of a mesh. In addition, information regarding to which meshes a site belongs was added to the site details pages. This is to help in understanding the mesh system that we have created. A new distribution of the central data collector software (for AMP) was made and added to the Web page. Also the license and NSF acknowledgment information was added to the Web page (it has always been in the software itself). A problem with the script that updates the AMP splash pages was fixed. A favicon for the AMP Web pages was created (a little meter that looks like a sun). A data banner (GIF image) with the new acknowledgment information (with both cooperative agreement numbers and begin dates) was created for use on the Web pages, especially those with data and/or tools. The MTM page (Meet the Team) was moved from the moat server to mave (which earlier in the period had been set up to be an additional Web server); it was also updated to reflect our staffing changes. Work began on the revising MNA front page, including rewriting the acknowledgement section. Also, numerous strategy and planning sessions were held re options for easing the process of making changes to the Web pages (SSIs, tunneling, Perl scripts), using templates and their structure and the templatization of the MNA files. A script was written to perform a task similar to Server Side Includes as we determined this was the cleanest way to obtain the functionality desired. A new project was begun to manage (and create) dynamic front page content. It has been extended to serve as a "rolling reports" system as well, whereby as items are updated for use on the dynamic front page, all information from the weekly reports will be archived and retrievable from a database. From the database, the archive page can be generated and the monthly reports can also be created in a draft form, in addition to the generation of the dynamic front page content. It is currently in the design development stage. Preliminary discussions, planning, and development of ideas regarding a new Web page highlighting our international collaborations were held. Some preliminary work was performed (directory structure, preparation of photographs, creation of the index page, a header banner, and other artwork). More detail on these activities can be found in the quarterly reports for this reporting period, available at: Papers, Publications, Presentations, and Conference/Meeting Participation~ Presentations and Meetings ~SC2002 - We installed an AMP monitor at the Supercomputing Conference at the Baltimore Convention Center (SC2002)in the SCinet NOC; it performed well during the conference. We also worked with Matt Zekauskas of Internet2 and Casey O'Leary of Pacific Northwest Labs at the Baltimore Convention Center to install an OC48 demonstration monitor that we provided for the SC2002 convention (at their request). Bud attended the conference as well as overseeing the installations of the AMP and PMA machines. Ronn Ritke attended and participated in a meeting with Greg Monaco (NSF), John Towns, and Jim Ferguson (both of NLANR/DAST) in Baltimore. APAN/PRAGMA Meeting - Ronn attended and presented as the first presentation in the Measurement Session. While there, he continued discussions regarding collaborations for both PMA and AMP activities with key network researchers throughout the Asian-Pacific region. IMW (Marseilles, France) - Jörg attended; while there he spoke with Matt Zekauskas regarding some of the plans on continuing to work on I2 measurements. They have their plans, which he also announced there, and it is very complementary. We should be looking at closer integration, the combined efforts appear to give a better thrust. He also announced the PAM2003 Workshop and distributed flyers. 10GigE conference (at SDSC)- Ronn presented and announced that NLANR/MNA now has an OC192 monitor platform in place, connected to the DTF OC192 link to Chicago. AMPATH meeting in Florida - Jim Hale represented NLANR/MNA at the meeting, giving a presentation on our current activities. While there he met with Tony Rimovski of NCSA, John Towns of NCSA, Julio Ibarra of Florida International University, Maxine Brown of Star Tap, and Thomas DeFanti of Star Tap. Fall I2 Member Meeting - Ronn attended and presented. NSF PI meeting in Virginia - Ronn attended and was on the Measurement Panel. Australian UNIX Users Group Annual Winter Conference in Melbourne (AUUG2002) - Jörg presented his paper with Ian Graham and two students from the University of Waikato Comp. Science on duplicated packets in an IP trace taken in 2001 at the University of Auckland access link. Boulder Joint Techs meeting - Ronn met and spoke with a number of people before and after the meeting. Additional meetings were added on to the Joint Techs Meeting. Liaisoned with and updated Tom Greene from NSF who also attended. IST SCAMPI meeting in Prague - Jörg attended. SCAMPI is the European counterpart of HPC activities in the U.S., aiming at developing a 10 Gigabit network analysis platform (http://www.ist-scampi.org/). This coincides with our OC192MON activities. On an invited trip to Thailand and China, Ronn gave an invited presentation to the UniNet Group in Thailand. This was followed by three meetings and presentations in three cities in China. During the trip, multiple presentations, both formal and informal, were given regarding the NLANR/MNA activities, with special emphasis on collaboration and possible placement of AMP and/or PMA monitors. A report on this extended trip was submitted to NSF with an overview of possible resulting collaborations. Matthew Luckie gave a seminar to the Computer Science department at Waikato. He also gave a presentation on IPMP at the Waikato Univ. student meeting, resulting in much discussion. Both Tony and Jörg went to Auckland to attend a seminar and talk to Christophe Diot from Sprint about the measurement work he does on Sprint's network. While visiting AT&T Research Florham Park Jörg gave a talk on NLANR/MNA research work; it was well received. Chris Gross (new MNA undergraduate student researcher) prepared and gave a tutorial on PHP to a group which included the MNA staff as well as some members of the SDSC staff who work on Web pages (very well received). Ben Reesman researched the current state of the art in visualization of multidimensional data sets; he presented his findings internally. With Ian Graham, Jörg has made a preliminary proposal to have the Internet Measurement Conference 2004 (it will be upgraded from IMW to IMC next year) in Hamilton (NZ). ~ Papers and Publications ~ A new issue of the Network Analysis Times (NATimes 3.2) was published (with an international collaboration theme). Print copies were distributed widely, including at SC2002 in Baltimore, the IETF November meeting in Atlanta, the NSF PIs January meeting in Virginia, and throughout SDSC. The look of this issue greatly benefited from the work of Ben, the new student whose area is images and the development of Cichlid. http://moat.nlanr.net/NATimes/NAT.3.2.pdf "NLANR's Active Measurement Program: Network Knowledge Leads to Practical Payoffs" was published in Online, Vol. 7 (3), February 5, 2003. Online is the SDSC/NPACI biweekly newsletter. http://www.npaci.edu/online/v7.3/nlanr.amp.html Matthew registered and submitted a paper to Sigcomm 2003: "Using IPMP to Identify a Bottleneck Router" which discussed some of the bandwidth estimation work with IPMP on the CRCnet wireless network. Jörg wrote and submitted to PAM2003 a paper on performance assessment of passive Internet monitors during the weekend of December 6 and the following Monday (NZ time), with coauthors from Endace. Todd Hansen wrote an abstract for PAM2003 submission for a paper about applying active measurement techniques to mobile satellite links. Klaus Mochalski (Univ. of Leipzig) has drafted a paper on delay measurements; it is going to be a joint paper with PMA. An article on the recent PRAGMA2 Meeting presentation in Korea was completed and submitted for publication in the Korea Institute of Science and Technology Information (KISTI) Supercomputing Newsletter. The NLANR/MNA STI award press release was written, and after review by NSF, released. http://www.npaci.edu/online/v6.16/nlanr.agmt.html More detail on these activities can be found in the quarterly reports for this reporting period, available at: Collaborations and Student InvolvementWe 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 HPC 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 high performance community (HPC), 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 organizations and researchers.
In August 2002, Ronn travelled to Thailand and China by invitation; several potential collaborative efforts were explored and/or begun.
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. Jörg traveled to Europe this period and had many positive discussion regarding potential collaborations:
We also had discussions with many other groups from all over the world regarding potential collaborations:
Active collaborations with HPWREN and ROADNet as can be seen in the measurement and analysis work being done on the wireless network (HPWREN) by students of those two groups. Received approval for the new set of four PMA monitors in Korea, Germany and Norway, on the plate is to contact the sites and negotiate on shipment and installation. Hans-Werner Braun met with Joel Apisdorf, he was the original developer of the OC3 monitors upon which we based the initial NLANR MOAT work. We are working on issues related to IBM disk reliability with Gordon Hughes and the Center for Magnetic Recording Research (CMRR) on campus (UCSD). CMRR conducts studies and research into many areas of magnetic recording. CMRR has conducted major hard disk reliability studies, some in collaboration with IBM. Gordon is working with an IBM storage division vice president, Roger Hoyt on some collaborative efforts. Gordon is very interested in our issues and has consented to take some steps toward opening a study with IBM related to the use we are making of the IBM disks. (This activity was brought about by the high number of disk failures that both AMP and HPWREN were experiencing.) We continued to run the NLANR Traces User Community regarding the needs of the passive measurement community. http://moat.nlanr.net/PMA/traces.html Through the strong concerted efforts of several members of our team, significant progress was made towards our hosting of the PAM2003 Passive and Active Measurement Workshop: registration was opened, submitted papers were assigned, reviewed, accepted (or rejected), and authors notified, and the PAM2003 Web pages continued to be developed. The quality of papers submitted for PAM2003 was very strong this year and will likely result in an excellent workshop. In addition, a number of students submitted papers and we will give some support to student authors to assist in their participating. AMP - Received a couple of European AMP requests (Switzerland and Amsterdam) as a result of Jörg getting in touch with his contacts. We are also considering an AMP in Mexico (CUDI). Discussions continue. Several inquiries for PMA trace were received, including from UPC in Catalonia. ~ ~ ~ Students Matthew Luckie, an NLANR/MNA researcher and graduate student (Computer Sciences PH.D. program, University of Waikato), has the lead on two significant AMP projects, IPMP and IPv6. He is responsible for the development, implementation, and measurements for these major efforts. Matthew is enjoying the close collaboration on IPv6 with Bill Owens (NYSERNET) and Joe St. Sauver (U. Oregon). He has also created Scamper, an IPv6 version of CAIDA's Skitter; it traceroutes in a parallel method to a list of IPv4 and IPv6 addresses. He is collaborating with CAIDA on this project. New student researcher Chris Gross began work as the new PMA student researcher the last week of October. Chris is a first year student at UCSD. He came on board to support and help in advancing the PMA project. Chris began his work with the development of new metrics for the project. He was assigned the task of helping to develop an online payment system for this year's PAM conference, which was successfully completed. After which, he resumed his work on metrics and began work with Jörg to develop and create a real-time PMA analysis infrastructure. Ben Reesman accepted the second student researcher position and began work in early November. He is predominantly working on the Cichlid 3-D Visualization System, the development of which has been dormant since Jeff Brown graduated and left the group to enter the UCSD Computer Science Ph.D. program. Ben is also a first year student at UCSD. He also performs other activities in support of the group (other image work, etc.). He has been working extensively with Cichlid in an effort to modernize and mature its current implementation. Extensions for both efficient connected stream and reliable unconnected datagram communications, as well as support for real-time data generation are in progress. Cichlid is being reimplemented in C++ in order to accommodate the addition of graph types and to leverage cross platform GUI toolkits. A GUI is being designed for Cichlid which will hopefully be completely cross-platform between Windows and Unix/X11 platforms. We are looking forward to working with Chris and Ben over a very productive number of years (hopefully four!). More detail on these activities can be found in the quarterly reports for this reporting period, available at: Other ActivitiesThis period had a quite a bit of activity with regard to bringing on additional staff, primarily regarding the AMP FTE (full-time employee) Software Engineer opening (which is in regard to the reimplementation of the AMP software). We also posted the AMP/PMA Assistant Systems Administrator (as an FTE) and openings for student researchers. As part of the interview and selection process for the AMP Software Engineer opening, a coding skills test was administered to each of the candidates (Perl and C problems of varying difficulty). The programming test turned out to be a very important and useful part of the interview process and we have decided to continue to use that approach for all positions, including the student positions. A large amount of time and effort was spent by several members of the team on staffing activities. Jim Hale became a full-time UCSD/SDSC employee (50% NLANR/MNA, 50% HPWREN), as the AMP/PMA assistant systems administrator. Cooper Nelson started half time with MNA in mid-September (having previously worked for HPWREN in combination with Scripps Institute of Oceanography), he will be assuming a technical support role for both the PMA and AMP projects. David Cheney, having graduated from UCSD in June 2002, was rehired on a part-time basis to provide assistance to the new student(s) and work with Maureen on various Web projects (writing the back-end scripts), including continuing the work on creating a system and related scripts to ease changes to the MNA Web pages (for now and in the future). Two new students joined the staff, Chris Gross and Ben Reesman; both are first year UCSD students. We had an outstanding group of student applicants. In addition to the superb three finalists, there were a number of other students who stood out for various reasons and we will be keeping (and sharing with HPWREN/ROADNet) their resumes, etc. Shared some of the best candidates with an SIO research group associated with HPWREN/ROADNet (they chose one of the students). We also brought one of the students to the attention of Ian Graham of WAND (Waikato). AMP Project Lead, Tony McGregor, began his increased time (to 80%) first week of February 2003. He will be doing the reimplementation/repackaging of the AMP software. More detail on these activities can be found in the quarterly reports for this reporting period, available at: Program PlanPlease see: http://moat.nlanr.net/Reports/MNA/PP2003-2004.html - 30 - Back to the Top |
