NLANR/MNA logo

 

NLANR/MNA Activities Summary - November and December 2002

Highlights

Good progress was made this period towards a number of our goals, including the following:

  • NLANR/MNA now has an OC192 monitor platform in place and connected to the DTF OC192 link to Chicago (located at SDSC).
  • Working on a new testing architecture that allows us to add new tests, and control which monitor does which tests from the database.  We have gotten to the point where it's hard to manage the different tests we're doing with the ad hoc approach that we currently have on the AMP mesh. A proposal (guidelines) for how we can extend the AMP testing architecture to accommodate these needs has been written, and is being as a guide to implementation.  
  • Taking and analyzing measurements with IPMP on the wireless network with an aim of measuring the bottleneck on the CRCnet (Connecting Remote Communities Network) wireless network which is located in a rural area west of Hamilton, NZ (near Univ. of Waikato).  In order to achieve this, a Linux implementation of IPMP was created so that it could be deployed on CRCnet.  
  • Began analysis of IPv6 data over the 6 week period during which we have been collecting data.  Also AMP IPv6:IPv4 comparisons can now be generated upon  request vs. the previous generation of static pages every 15 minutes.
  • A new issue of the Network Analysis Times (NATimes 3.2) was published (international collaboration theme);  print copies were distributed widely, including to SC2002 in Baltimore and the IETF November meeting in Atlanta.  
  • Both an AMP and a PMA machine (OC48) were demonstrated at SC2002 in the SCinet NOC.  
  • 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 1-2 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.  

Passive Measurement and Analysis (PMA)

  • Successfully acquired and implemented a platform for a OC192 monitor supporting an Endace Dag6 card. NLANR/MNA now has an OC192 monitor platform in place, connected to the DTF OC192 link to Chicago (located at SDSC). (Done in record time, a bit of a thrash, but successful all the same.) The monitor was installed and the tap put in place. Although a lack of 10GigE traffic left us with nothing to monitor. [Bud Hale, Jim Hale]
  • Ian Graham and David Miller started testing the OC192MON system in Burlingame, the results were astonishing. The system could keep up with a million cells per second and take packet sizes of up to 280 bytes, which is equivalent to 3.2 Gbits/sec, which is presently the limit with the implementation. The system has several minor flaws, the start/stop backpressure implementation falls over, which makes the system unusable if the software can't keep up processing.  VHDL folks in Hamilton are working on this matter. Stephen Donnelly has written a real-time test program, bypassing the dagsnap tool, to lift the present bandwidth limits for test.  Testing continued at SDSC, Bud had organized with Jay Dombrowski's group to get a 50:50 tap into the link from SDSC to Chicago. A pity there are no systems connected to the link at present, so no traces could be taken and tests be run. From a distance I understand that everyone was impressed and pleased with how well things went. Great job by Bud and Jim Hale, especially. [Jörg Micheel]
  • On the OC192MON front folks have found the last bug within the FIFO, it was a problem with reset on startup, and there are no known residual problems, the production boards will go ahead shortly, for availability around January. We might want to spend some time thinking where we want to place an NLANR OC192MON. Perhaps Abilene or StarLight might be good choices, but starting off with SDSC could be better. Will be traveling in January and plan a field trial of an OC192MON.
    [Jörg Micheel]
  • Had a phone call with Chris and took him on a introductory tour of PMA. Then in late November took about an hour to talk to Chris Gross on his progress with initial trace analysis. Looks like Chris is now familiar with how to do deal with regular Dag trace files. We discussed what we want to do in order to proceed with some interesting new analysis work which will involve permanent monitoring and a low-volume data stream to be displayed as RRD graphs on the pma data server. Task for Chris is to come up with a set of passive measurement metrics, as well as priorities, to tackle within this project. We are going to start off with using the new SDA (SDSC Abilene) monitor, as it is both convenient, and also allows us to get some feedback from ENS, Jay Dombrowski's group. [Jörg Micheel]
  • Started thinking about metrics, their meaning, and how to possibly measure them, sent the most meaningful ones to Jörg for feedback.  [Chris Gross]  Reply:  have received an update from Chris regarding the metrics we want to pursue with PMA real-time analysis, haven't had enough time to reply yet, but it looks like an interesting discussion. [Jörg Micheel]
  • Worked with advancing a program that extracts information from traces,  trying to see what types of data traces can offer for analysis.  Discussed organizing a development environment for our PMA projects with Cooper.  [Chris Gross]
  • Created a CGI driven stats analysis of pma.nlanr.net (Web log). Writing script that will merge http and ftp logs to display the results in one report. Worked on scripts to merge ftp and http logs of trace downloads (ftp to apache log conversion). Wrote a script to convert the ftp logs into a format awstats can read.  [Cooper Nelson]
  • Moved my development environment to pma, all scripts working in a production environment. Scripts in place to automatically generate graphs of http traffic and site traces.  Worked with on modifying Jörg's code to draw independent traces for each PMA site.  Began working with Jörg's C log analyzer. [Cooper Nelson]
  • Configured and shipped the OC48 PMA machine for Supercomputing 2002 [Jim Hale]
  • The OC48 monitor deployed at the SC2002 convention in Baltimore MD has been returned.  It has been checked for shipping damage and/or functionality problems and reshipped to Jörg in New Zealand for some testing.  [Bud Hale, Jim Hale]
  • Continued work on my sampling research:  Continued to make progress on my sampling analysis scripts. I edited a couple of my strips for clarity, and better style.  Reading up on other ways of graphing the data that I've been working with.  Fixing a bug in the cron jobs that run my sampling scripts.  Starting up on an abstract for my sampling work. Fixed a few more of the bugs effecting my numerical analysis.  Finally got my new graphs working.  The Web page still needs some major cosmetic changes.  These new graphs plot the accuracy of sampling for specific ports, as a function of the size of the data coming through for that port. moat.nlanr.net/~jfields/sampling.html  [Justin Fields]
  • Started a very interesting discussion on how to get a handle on the current state of deployments.  Tony also brought in some valuable comments and it appears we have a strategy on how to make progress with this.  [Jörg Micheel, Hans-Werner Braun, Ronn Ritke]
  • PMA existing sites, maintenance and troubleshooting:  
    • Old Dominion U. resolved their router problems and also put the passive OC3mon back on line after completing a router reconfiguration and network redesign. It should again be reporting data soon.  [Bud Hale]
    • In the process of final testing of the TAU and FRG machines.  [Bud Hale, Jim Hale]
    • The OC48 monitor for the Mid-Atlantic Crossroads (MAX) GigaPop was rebuilt in late November.  Then when it appeared ready to ship, it seems to have turned up a SCSI cable problem at the last moment. Will be return shipped when ready.  Resolved some issues with Dan Magorian.  [Bud Hale, Jim Hale]
    • Rice U. - Worked with site people to develop more support at the Rice location as well as the Texas GigaPop. It appears more support will be available in the near future. I was able to get the OC3 ATM monitor at Rice back up; but still need help at the Texas GigaPop. The site technician, Stan Barber has been on extended travel and leave and not available.   [Bud Hale]
    • Ohio State continues to have a problem requiring frequent rebooting. [Bud Hale, Jörg Micheel]
    • The APN site seems to need a reboot. However due to the holiday season the staff seems to be at skeleton level causing problems getting help on site. [Bud Hale]
    • The Advanced Network OC12mon has gone off-line. I expect that is due to the shutdown of the I2 network connection at the Armonk, New York location. I sent a message to Matt Zekauskas to that affect and indicated I expected the machine to be returned to SDSC for preparation for another I2 connection at Ann Arbor, Michigan.  [Bud Hale]
    • The IP address conflict existing with the PMA monitor on the Abilene network in the SDSC machine room has been resolved. There remains an name error in the SDSC name server but that will be resolved as soon as Tom Hutton is available.  [Bud Hale]
    • Inclusive list of sysadmin tasks

Active Measurement Project (AMP)

  • Amp has developed to the point where it's hard to manage the different tests we're doing with the ad hoc approach that we currently use. We have developed a proposal for how we can extend the AMP testing architecture to accommodate these needs.  Planning to use this as a guide to implementation in the coming month(s).  Some progress has already been made, but there's a lot more work to be done.  The Web page and database to create new tests and most of the one to assign monitors to tests are complete.  An important detail I didn't include in the design document completed earlier, is how to handle defaults.  I've decided that it's 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.  Otherwise, I'm following the design document. [Tony McGregor]
  • Created an international mesh with all the non-US sites in it and also sdsc, psc and bu. Over time, I'd like to migrate the international sites out of the main HPC mesh. [Tony McGregor]
  • Moved the four AMP sites that had been in the HPWREN mesh back into the main directory.  There was no practical problem with them being separate, but it didn't make any sense and was confusing at times. We may have to remake a group in the future when we reintroduce HPWREN machines but in the meantime it's simpler this way. [Tony McGregor]
  • Worked on fixing the on-demand throughput tests. Installed the 3.0 libc.so to get the old binaries to run.   [Tony McGregor, Matthew Luckie]

    ~ ~ ~ IPMP
  • Major accomplishment was taking and analyzing measurements with IPMP on the CRCnet wireless network with an aim of measuring the bottleneck. The dispersion graph (http://voodoo.cs.waikato.ac.nz/~mjl12/dispersion.png) is very interesting - there is a big gap between 3 and 5 milliseconds in the middle of the network around WTU that is caused by contention on the MCG <-> HSK and HSK <-> WTU half duplex links.  Because we are sending two packets, there will be contention at WTU when the first packet is turned around while the second packet is still in transit.   [Matthew Luckie]
    • While at SDSC worked to get some measurements with IPMP on the CRCnet (Connecting Remote Communities) wireless network.     [Matthew Luckie, Tony McGregor]
    • Wrote a script to break the round-trip-time on the wireless network into the delay contributed by each link on the network.  I found that the Linux clock is not of sufficient resolution to get a meaningful breakdown, so I'm looking at the PPSKit patch to provide me with a nanokernel. [Matthew Luckie]
    • Patched linux-2.4.19 with the PPSkit to give my IPMP kernel a better time source; it seems to be working a lot more sensibly now.  Ordered the relevant cable to work with the Trimble Acutime so that we'll have a decent PPS source shortly. [Matthew Luckie]
    • Modified ipmp_ping to accept a count parameter so that I could do packet-pairs.  Hoping that when the PPSKit patch goes out that I'll be able to identify the bottleneck on the wireless network with a two-packet exchange.  You don't need synchronised clocks to do this. [Matthew Luckie]
    • Working on the Linux implementation of IPMP.  The ipmp_ping client is at a point where it is able to deployed on CRCnet now.  I had to do a fair amount of work to get the ping client to work under Linux; for example Linux does not yet have a getifaddrs implementation, so I needed to acquire a list of local IP addresses elsewhere. [Matthew Luckie]
    • I wrote a bit more IPMP code so the user can define what the source address of the packet will be.  I found a problem with some of my linux kernel code when it interacted with raw sockets.  I fixed that and then gave the code to some CRCnet people to deploy.   [Matthew Luckie]
    • Wrote code to make the linux NAT work reliably with IPMP. Implemented stateful NAT / filtering of IPMP packets on linux. Modified the userland and kernel code for BSD to work with these modifications.  In the process I re-coded most of the BSD kernel code to do incremental updates to checksums and made the code significantly tighter (I removed at least 100 lines of kernel code). Started work to backport IPMP to previous versions of linux so that it is more useful on CRCnet. [Matthew Luckie]
    • Made the IPMPlinux code timestamp echo responses "on the way up the stack" by adding my own code to do raw socket interaction with IPMP sockets.  The linux "stack" is horrible like that. Implemented an ioctl to send a time-stamped IPMP echo request from kernel space in linux.  Ported measurement daemon to work in linux. [Matthew Luckie]
    • Created a patch to linux-2.4.19 and put it up at http://mave.nlanr.net/~mjl/linux-2.4.19.ipmp.patch Friends with linux boxes have tested the code and found some problems, working to sort those out. [Matthew Luckie]
  • Writing some scripts to extract the per-link RTT from an echo packet. Won't be able to break down the delays as much as I would have liked re asymmetric paths.  Might look at using a third-party graphics library to draw the topology as seen by IPMP when we have data available from all ends of the network. [Matthew Luckie]
  • Spoke with some hardware gurus at Waikato about designing and implementing an IPMP dongle that has a male and a female Ethernet connector and some VHDL to insert path records on any IPMP echo packets that it sees.  I've got a loan on some VHDL books to have a read over the holidays. [Matthew Luckie]
  • In early November, there was quite some activity  on the IPMP proposal. Jon Bennett, from Motorola, released an updated version of the draft.  Sadly, he chose to do it without consulting any of us, which is a pity because we would be better working together. Various people (including one of the IETF IPPM working group chairs) have written to me asking what was going on.  There are quite a few contentious issues in his draft (as well as some obviously good improvements). I'm hoping to convince Jon to work with us, rather than taking the competitive approach he has. [Tony McGregor]
  • Spent a fair bit of time going over Jon Bennett's IPMP draft and mailing him and Matt Zekauskas about things.  Found out with too little notice to attend, that IPMP is on the IPPM agenda for the IETF November meeting. I would have tried to go, given a little more notice. Matthew and I have sent numerous detailed emails to Jon Bennett re specific issues.  [Tony McGregor]
  • Invited Jon Bennett to come during our visit to San Diego in December (at our expense). He accepted and we had discussions with him about IPMP.  [Tony McGregor and Matthew Luckie]

    ~ ~ ~ IPv6
  • Began analysis of IPv6 data over the 6 week period during which we have been collecting data.  Had a look at some interesting IPv6 cases by generating a bunch of graphs that showed how the distribution of IPv6 RTTs changed over time.  By looking for sudden shifts, I found the following rather interesting cases:  http://mave.nlanr.net/~mjl/interesting-ipv6/amp-gatech.amp-wisc.102.10.21.html http://mave.nlanr.net/~mjl/interesting-ipv6/amp-gatech.amp-wisc.102.11.21.html http://mave.nlanr.net/~mjl/interesting-ipv6/amp-aarn.amp-gatech.102.11.16.html  [Matthew Luckie]
  • Wrote three cgi scripts to do the daily comparisons for IPv6/IPv4 data sets dynamically, (gave them to Tony to look over, then put in cgi-bin). These dynamically generate the IPv6:IPv4 comparisons when requested, instead of generating static pages every 15 minutes. http://watt.nlanr.net/active/cgi-bin/v6_portal.cgi  [Matthew Luckie]
  • Worked with Joe St Sauver to get amp-uoregon back onto a v6 network; it is collecting v6 data again. His monitor had not been working with IPv6 for a while as something changed on the U Oregon network. [Matthew Luckie]
  • Got the IPv6 mesh going again after Tony changed the way tests are run from the amplets.  Put the individual crontab files for specific sites in the system manager so it won't happen again. [Matthew Luckie]
  • Got amp-korea to a state where it could be upgraded to 4.6 like the rest of the amplets. [Matthew Luckie]
  • Had some discussions with Warren Matthews of Stanford Linear Accelerator Lab (SLAC) on participating the IPv6 monitoring program. He has indicted that his security people will require that a second monitor be installed on the IPv6 network. It seems there is a security issue there related to both the IPv4 and IPv6 networks interfaces sharing the same box. It seems there is fear that one might appear as a back door for the other. Needs more review. [Bud Hale]
    ~ ~ ~
  • The AMP server data disks were reaching the capacity limit so the doarchive perl program that has been in use for more than a year was started; it failed. Tracked down a problem with the archiver  which turned out to be that the HPSS has a new version of FTP that behaves slightly differently.  The new software doesn't include quotes around path names. This caused the perl ftp library to not return the pathname from a pwd command. Was able to get the archiving perl script working with the new version FTP recently installed on the HPSS. Put together a custom version of the Net::FTP library that works with the new HPSS software and ran an archive. [Tony McGregor]
  • Full archive run completed on AMP server; it brought the disk fill down to 64 percent. Remaining task was to move these developed changes to the doarchive perl script on the VOLT server. The VOLT server disk configuration is different from the AMP server, therefore the doarchive script will not work as is on VOLT. Changes are needed to handle symlinks on VOLT that are not needed on AMP.  [Bud Hale]
  • Fixed the archiving program.  Converted the doarchive script so it would work with both amp and volt and ran it on volt. Still had problems; tracked the problem down to the HPSS occasionally returning errors.  In the process, discovered a bad directory on AMP, which Bud fixed with an fsck.  [Tony McGregor]
  • Worked on the doarchive problem; discovered that the main problem stems from intermittent failures of the HPSS and some faulty error handling code in the script.  In a nutshell, if HPSS drops an ftp session the script will attempt to reconnect once; if it fails it will die outright. [Cooper Nelson]
  • Fixed the HPSS problem by making the archive recover from the error and continue on.  Also fixed the code to make it start from where it got up to on the last run if it crashes or is killed.  When the archiver finally ran, it reduced the disk space to 66% full. [Tony McGregor]
  • Worked with the HPSS people re upgrading our HPSS data storage/archiving processes, both for AMP and for PMA. There are some questions about the best interfacing scheme to use with the HPSS. Discussions with the HPSS people have indicated that the SDSC HPSS predominantly supports two interface schemes. Those are PFTP (parallel file transfer protocol) and HSI (Hierarchical Storage Interface). NFS is not supported at SDSC and the API (Application Program Interface) is supportable for selected applications but not recommended for the NLANR/MNA archiving.  Also it appears our perl scripts can be easily converted to the HSI.  But the SDSC HPSS staff reports that shell scripts are used mostly with the HSI interface. This method is unique to the HPSS and was developed by the HPSS community. It consists of a UNIX shell script on user machines that handle all the transactions and data transfers with the HPSS. I will be working with that script to gain familiarity with it. [Bud Hale]
  • Reading about the HSI file transfer mechanism, which is what the HPSS folks recommend we use for AMP.  [Cooper Nelson]
  • Installed and distributed a new kernel to the amplets that supports connecting the keyboard after boot. Did some work with Matthew to restart the IPv6 updates with the new kernel. [Tony McGregor, Matthew Luckie]
    ~ ~ ~
  • Installed FreeBSD 4.7 operating system from ftp on the Ohm server again after an IBM system disk crash. Installed user accounts; had problems with transferring the contents of each account to the same folders on the ohm machine.  [Jim Hale]
  • As a result of interest in IPv6 accessibility and the stability of Apache1.3, Apache1.3 + IPv6 was chosen to be installed on the Ohm server. At this time the start command has not been performed till the configuration has been completed and the Web pages have been transferred from the AMP server.  [Jim Hale]
  • Input the new configuration into the web server # /usr/local/etc/apache/httpd.conf file, based on the patch pages Matthew created and recommendations Todd made.  [Jim Hale]
  • Discussed some unusual entries in the security log on the ohm server with Todd.  [Jim Hale]
    ~ ~ ~
  • Changed most of amp/volt and the amplets to process mail independently of the main NLANR domain.  From time to time things go wrong in the amp system and a lot of mail is generated because there are a lot of systems involved.  That has sometimes cause problems for the central mail system.  I may have missed some things (mail is sent for many different reasons, some of which are not immediately obvious) but will change those over as we find them. [Tony McGregor]
  • Conferred regarding AMP sites being able to create excessive amounts of email traffic under certain conditions. This discussion resulted in a plan to let that email traffic stay on the amp server (short-term resolution).  [Tony McGregor, Bud Hale]
  • Changed the email setup on the amplets. It turned out, for a fairly subtle reason, the newaliases command hadn't been run when I did the update last week.  Matthew did some work on the sendmail setup so that we wouldn't get mail from cron via moat anymore. I tested it all out and updated the amplets.  I'm pretty confident that we won't get mail, with the possible exception of the machines that were down. [Tony McGregor, Matthew Luckie]
  • The mail update revealed some problems with the way one of the sites had been setup in the system manager.  We may need to tighten our procedures there, or automate the process. [Tony McGregor]
  • Performed the OS upgrade on the calorie server. Prepared new kernel source for calorie based on the MSQL_DB that was on calorie before the upgrade.  The upgrade from BSD version 3.3 to 4.7 caused problems with the calorie server.  Advisories began coming in that calorie was not taking requests for new monitors. Consulted Todd and Matthew about the problem with the site request form, then fixed it.  With sendmail fixed, requests submittal is working well again.  [Jim Hale]
  • Discussions during Tony and Matt's visit led to a decision to rework the AMP system disk such that the kernel will support the 3Com GigE NIC, the Intel 10/100 Pro and the RealTek 10/100. [Tony McGregor, Bud Hale, Matthew Luckie]
  • Continued testing and conferring with Tony on the condition of the AMP site master disk. Have finished all testing on the image and it is complete. Together with some procedural planning the site image master disk is finished. It was successfully used to create the AMP site monitor for the Supercomputing 2002 AMP monitor.  [Bud Hale]
  • Went over the new master disk that Bud has created and cleaned up a bunch of stuff on it.  I created an empty meshes file on the master disk which is extra protection against the machine doing tests before it's been configured, or testing to the wrong sites while the configuration is happening. [Tony McGregor]
  • I finally figured out what was going on with the msql access control list and have the traffic stats being collected into the new database from all the machines (and the Web pages changed to use it).  It turned out that only the first matching database rule was being considered in the acl, not the most specific one. Removed the traffic statics from the main database and put them in a database of their own.  The stats are 13Mb compared with 400kb for the rest of the database.  Considering it is backed up each night moving them to a separate database seemed like a good idea. Hans-Werner truncated the backup file (on moat) for me. [Tony McGregor]
  • Met and worked with Tony and Matthew for the first time in person. Had some great instruction time with them.   [Jim Hale]
  • A new AMP monitor request was received from the HEAnet in Dublin, Ireland.  [Bud Hale]
  • Prepared and shipped AMP machine to Supercomputing 2002 [Jim Hale]
  • AMP existing remote sites maintenance and troubleshooting:  
    • U. of Cal., Berkeley (amp-ucb) - hit with a DDos attack that actually wedged an OC12 interface. The AMP monitor was out for about 12 hours before their router problems were corrected.  [Bud Hale]
    • Three copies of the new version AMP system disks have been shipped to sites for replacement of failed disks. Those sites are amp-apantyo (Tokyo), amp-aarn (Australia) and amp-csupomona (Cal. State, Pomona).  [Bud Hale]
    • amp-apantyo (Tokyo) - replacement disk has been installed, the site was restarted and is now collecting data again.  [Bud Hale]
    • Cal. State, Pomona  - still out due to a bad system disk. The replacement disk has been on site for awhile but the technician has been unable to install it. That should happen immediately after the holidays.  [Bud Hale]
    • UCSB hadn't been entered into the HPC mesh in the database after the upgrade to FreeBSD4.6; fixed that.   [Tony McGregor]
    • amp-nwu (Northwestern U.) - experiencing an outage
    • amp-uah (U. of Alabama at Huntsville)- fully reachable but is not transferring data.  (It's collecting measurements, but they're not being transferred to amp and volt.) Could not be started with a run of the system manager.   [Bud Hale, Tony McGregor]
    • amp-bcm (Baylor College of Med.) - remains unreachable with ssh. However it is still collecting data;; but cannot be updated (system manager control issue).  [Bud Hale]
    • amp-ou (Oklahoma U) - issue with the mesh assignment. That site also appears otherwise okay. [Bud Hale]
    • Old Dominion U - After an outage of several weeks due to site router issues (router reconfiguration and network redesign completed at the site), the site is back online.  [Bud Hale]
    • U. of Cincinnati  - had a short outage due to some local network issues but were resolved.  
    • U. of Alabama at Huntsville - has corrected a router blockage of ICMP packets, allowing pings to the AMP monitor. However there is still a problem with data transfers, under investigation.  [Bud Hale]
    • Fermi Nat. Labs. (amp-fnal) - had an outage of about two days due to a failed switch port which has been corrected.  [Bud Hale]
    • Mississippi State (amp-msstate)- is transferring data, however there is an issue of system manager control. To be resolved shortly. [Bud Hale]
    • U. of Central Florida  - issue seems to be resolved. The site has relocated the AMP monitor on an internal network and the IP pointing to the internal network. Will monitor it. [Bud Hale]
    • Three of the remaining AMP sites not yet upgraded to FreeBSD4.6 were upgraded, amp-wpi, amp-uwyo, and amp-okstate. amp-okstate has been restarted with system manager and is collecting data.  The two other sites, amp-wpi and amp-uwyo, are currently not accepting ssh-agent connections.  amp-korea has now also been upgraded.  [Bud Hale, Jim Hale]
    • uwyo (U. Wyoming) and wpi (Worcester Polytechnic Inst.) - There was a problem with the permission on the meshes file that meant when came back online they didn't collect data; fixed that as well as the program that generates those files from the database. [Tony McGregor]
    • Earlier, wpi (Worcester Polytechnic Inst.) was failing due to a problem with the Juniper router that was blocking port 22 even though it was not supposed to happen (discovered by the tech on site).   [Bud Hale]
    • amp-harv (Harvard) - prepared for moving to the new location. Currently in a transitional state due to the process of changing network connections.   [Jim Hale, Bud Hale]
    • Fixed some on-site network problems amp-clemson.  [Bud Hale]
    • Made contact with Eric Goodrich of Harvard University to set aside time to work on getting amp-harv back on line after the machine is moved to a new network. [Jim Hale]
    • Several sites came back online this period, did a system manager run on them.   [Tony McGregor]
    • Inclusive list of sysadmin tasks

Additional Performance Measurement and Analysis Activities

  • Network utilization script:  working on getting options to limit the number of results from each router or interface and to generate extended results for each host. These options may or may not sound very interesting by themselves, but when combined it allows for a breakdown of each router and the peak traffic on each one, with a link down the page to information about activity on each router. This seems like a good overview of peak traffic on each part of the network. Also did a successful one year long query:  http://stat.hpwren.ucsd.edu/~wgahr/results.html [Bill Gahr]
  • Tested, debugged and fixed all of the Lucent MIB scripts.  Obtained an archive copy of 2001 Lucent datafile for debugging purposes. Finished writing a Perl subroutine to extract data for Tsunami continuously and did research on time series analysis.  [Zhao Li]
  • Completed a document outlining Web interface guidelines to be used on stat (the measurement and analysis pages of HPWREN).  [Jeff Baker]
  • Updated the Router Traffic Data on stat. Began work on updating the Switch Traffic Data page, staying more or less within accordance of my newly completed Web Interface guidelines. Updating the cgi pages in the MIB Data portion of stat. Touching up a backup version of the UPS Data page. [Jeff Baker]
  • Updated the Lucent and Tsunami MIB Data page, wrote text for the Router Traffic Data page.  [Jeff Baker]
  • Finalizing the article for submission to ACM Crossroads. [Jeff Baker]
  • Created a new version of the HPWREN topology map. http://hpwren.ucsd.edu/topo.html   [Hans-Werner Braun]
  • Replaced the broken PAM machine <a combination passive and active measurement monitor used on the HPWREN network>. [Hans-Werner Braun, Jim Hale]

Documentation, Networked Data, and Tools

  • Spent some time on the traceroute graphing code to make it possible to generate the image map separately, which is useful for the CGI scripts I have, and to send the image itself to stdout if requested.  I have some bugs on that code which need attention over the coming weeks. [Matthew Luckie]
  • Added a Web page to list the sites (AMP) that are a members of a mesh and added the meshes a site is in to the site details pages.  Bud had suggested this will help people to understand about the meshes. [Tony McGregor]
  • Discussions with Cooper on http/ftp download display. Cooper dug out a new piece of Web statistics package, awstat, which looks like a promising way of gaining control over this graphing issue.  Have given him directions on how to structure the various views and make a set of pictures of the past as well as daily scripts available to bring up a permanent display of activity. 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. This is good news, it reflects that, on average, all our data is useful, is being looked at and studied. [Jörg Micheel]
  • Created a Web server on mave, to support some of Maureen's needs. [Hans-Werner Braun]
  • Worked with Hans-Werner re getting a Web server onto mave, streamlining staff home pages and the /Drafts directory. H-W has the server up and operational and the MTM page now maps to the mave directory. I moved the image over, but will have to fix the navbar (put in full pointers); the staff home pages don't work since they're still on moat. [Maureen Curran]
  • Per HWB's suggestion that the MNA home page should be alive, fresh, and focused on our current accomplishments (vs. static), I redesigned and worked on the layout of a new MNA home page.  Redesigned the MNA Web page template, using the 680 pixel page layout, including the data banner; also added last updated line.  Completed the new directory, wrote a brief on explaining ANR, completed the design and layout, posted for Ronn's and HWB's comments.  [Maureen Curran]
  • Meeting regarding NLANR/MNA Web interface.  Afterwards, Ben gave a presentation on multidimensional visualizations. [Hans-Werner Braun, Maureen Curran, Todd Hansen]
  • Sorted out the new logos to use, decided on just two sizes , regular(474x60) and larger (600x76). Plan to use the 474 size for Web pages in general, and the 600 size for the home page.  Uploaded them to Images directory; sent an email to all re pointers, sizes, of both logos and the databanner, etc.  In the future, all pages will have the NLANR/MNA logo and the data banner as headers/footers respectively. [Maureen Curran]
  • Working with Dave to create a management system for creating and maintaining a section of dynamic content on the MNA home page and in doing so create a database which can be polled to create the monthly reports and an archive page of our activities (by subject, in chronological order). (This grew out of an idea of Todd's.)  [Maureen Curran]
  • While continuing to work on the redesign of the MNA home page, began working on the specifications for Dave to use in writing the appropriate scripts and designing the back-end of the system. [Maureen Curran]
  • Did a quick redesign of the NATimes home page when I posted the PDF of the new issue (including the new cooperative agreement numbers, etc.).  [Maureen Curran]
  • Continued work on more of the International Collaborations Web pages.  [Jim Hale]
  • Meeting to review the International Collaborations Web pages work in progress.  [Ronn Ritke, Jim Hale, Maureen Curran]
  • Provided Jim the acknowledgment text for the IC Web page and wrote a paragraph for use on the people page under construction.  [Maureen Curran]
    ~ ~ ~ Cichlid 3-D visualization system, other work on images
  • Emailed Jeff Brown about Ben starting and that he'll be working on Cichlid (he replied that he'll be available to answer Ben's questions, etc.).  [Maureen Curran]
  • Learned about and worked with Cichlid; studied the Cichlid source code. Got it to compile and run on Windows.  Wrote a simple Cichlid server. Experimented with the tolerances on the Cichlid Server APIs (i.e., how big of a value is acceptable for this or that etc.)  [Ben Reesman]
  • In response to a request of Hans-Werner's to research the current state of the art in the rendering of multidimensional datasets, researched, read, and reviewed current techniques and papers, dissertations, and documentation on the Web regarding visualization.   [Ben Reesman]
  • For the presentation of the visualization review project:  developing and preparing images, preparing slides, and selecting the appropriate data to the relevant points.  [Ben Reesman]
  • Worked on animated visualization demos for the anemometer data that is collected by HPWREN. In the process of developing a workable 3-D representation that could be both accurate and aesthetically pleasing. [Ben Reesman]
  • Learning OpenGL; Experimented with different OpenGL drawing routines that could be used to implement some of the visualization strategies we would like to see in action.  [Ben Reesman]
  • Just started to work on a Cichlid server in C++, to explore the linkage possibilities.  [Ben Reesman]
  • Tried to create the basis for an implementation of the Star Coordinates system, but am meeting with little success. Hit-detection is somewhat complicated.  [Ben Reesman]
  • Other work creating or improving images for use on Web pages and various publications, including the 600 pixel MNA logo, a smaller version of the PAM2003 logo.  [Ben Reesman]

Papers, Publications, Presentations, and Conference/Meeting Participation

  • Completed and published the NATimes 3.2 issue (international collaboration theme):
    • Did final edits, including making authors' changes on half a dozen articles (a couple of which are very short); wrote a brief for CANARIE's AMP deployment for the issue.  Wrote new NSF acknowledgment section and copyright/subscription footers (since the natimes@nlanr.net email is disabled, changed to a pointer to the NATimes feedback form for comments, subscription requests, etc.).  Compiled and reviewed all the images, and made preliminary choices about which to use, developed some ideas for layout.  Arranged for printing and distributed printed copies of the new issue of the NATimes to SC2002 in Baltimore and the IETF November meeting in Atlanta. http://moat.nlanr.net/NATimes/NAT.3.2.pdf  [Maureen Curran]  
    • Layout and formatting of text and images for the NATimes issue.   [Ben Reesman, Maureen Curran]
    • Sent 20 printed copies of the current NATimes to Manhee Lee (KISTI/Korea) for an event of his. (He has an article on Kreonet in the issue, and is part of the front page collage.)  [Maureen Curran]
  • 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. The Baltimore OC48MON was needed as one of the machines to run comparative tests on. The other two machines under study are the Dell 1650 and 2650 (also used as OC192MON soon). [Jörg Micheel]
  • Working on the Internet draft for IPMP: updated IPMP draft with fields for ID / sequence number.  Removed data field which became redundant. [Matthew Luckie]
  • Wrote an abstract for PAM2003 submission for a paper about applying active measurement techniques to mobile satellite links. [Todd Hansen, ROADNet; Maureen Curran (ed.)]
  • Klaus Mochalski (UoLeipzig) drafted a paper on delay measurements, it is going to be a joint paper with PMA. Also need to provide more comments and changes on this one.  [Jörg Micheel]
  • Attended and presented at the Fall I2 Member Meeting.  [Ronn Ritke]
  • Attended the European 10GigE monitoring platform meeting, IST SCAMPI (late October). Met with a number of folks that I hadn't heard of before. It was quite interesting to see what everyone is doing, and some of the discussion reminded me of things that I had seen and experienced five years ago, aka discussions about file formats and the like. I presented our approach to passive monitoring, and it was (as far as I can tell) very well received by most of the folks there. The perception is that only with the Dag technology they are likely to get to where they want to be. I showed the prototype board, of course, and we discussed technology enhancements and roadmaps. We need to discuss how we want to create a research collaboration with these folks, there is quite some room and opportunity. [Jörg Micheel]
  • SC2002
    • Installed an AMP monitor at the Supercomputing Conference at the Baltimore Convention Center (SC2002); it performed well during the conference. [Bud Hale]
    • Sorted out the setup with the SC2002 machine and set it going.  Also, spent some time going over the issues with Bud. [Tony McGregor]
    • Worked with Matt Zekauskas of Internet2 and Casey O'Leary of Pacific Northwest Labs at the Baltimore Convention Center to install the OC48 demonstration monitor we provided for the SC2002 convention. After some troubleshooting with Jörg, we had it working. It was available to monitor the Pacific Northwest Labs OC48 connection into the conference. [Bud Hale]  
    • We were going to shoot for a capture session, I had all the scripts in place, but something went wrong and the capture did not start. Most of the time traffic was relatively low, with 800 packets outbound and a few thousand outbound, a mere few Megabits per second of load. [Jörg Micheel]
    • On-site coordination with Bud Hale at SC2002 in Baltimore regarding NLANR AMP and PMA units in SCinet.  [Mike Gannis]
    • Attended SC2002 and had a meeting with Greg Monaco, John Towns, and Jim Ferguson (both of NLANR/DAST) in Baltimore. Some set up work was done at SC2002 (NLANR partner signs, NATimes were put out at the NPACI booth, etc.). [Ronn Ritke]
    • Meetings with Rozeanne Steckler (SDSC Assoc. Director, Education) about the NSF-Thailand workshop. Talked about the meeting with Bill Chang at NSF in Virginia. Arranged to meet early Mon AM and take the train from Baltimore to NSF headquarters in Virginia and back again for SC2002.  Called Greg Monaco and spoke with him about activity updates and the NSF-Thailand Workshop. He gave a positive response to NLANR/MNA supporting that effort.  [Ronn Ritke]
  • Spoke with Matt Zekauskas at IMW (Marseilles, France) on 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. Managed to announce PAM2003 there, thanks to Maureen for managing to get the flyers here in time. [Jörg Micheel]
  • Presentation at the 10GigE conference (at SDSC), announced that NLANR/MNA has an OC192 monitor platform in place, connected to the DTF OC192 link to Chicago. [Ronn Ritke]
  • Prepared and gave a seminar to the computer science department at Waikato (had some excellent questions).  Also gave a presentation on IPMP at the Waikato Univ. student meeting, resulted in much discussion.    http://voodoo.cs.waikato.ac.nz/~mjl12/ [Matthew Luckie]
  • Prepared and gave a tutorial on PHP. Created a power point presentation to go over with the group <MNA staff as well as some members of the SDSC staff who work on Web pages>.  [Chris Gross]
  • Updated Ronn's set of NLANR slides for the AMPATH meeting presentation.  [Jim Hale, Ronn Ritke, Maureen Curran (ed.)]
  • Worked on preparation for the AMPATH meeting in Florida (late Jan) presentation; numerous iterations.  [Jim Hale, Ronn Ritke]
  • Ian Graham and I have made a preliminary proposal to have the Internet Measurement Conference 2004 (it will be upgraded from IMW to IMC next year) in Hamilton. [Jörg Micheel]

Collaborations and Student Involvement

  • Showed Bill Owens some of the interesting IPv6 cases I found (listed above).  He had explanations for some of them, but not the aarn->gatech on 16 Nov case.  That case is unusual to say the least. [Matthew Luckie]
  • Yan Chen from UCP asked for two months data. I've sent the first month and am awaiting the request for the second month.  They've written one paper about AMP data already.  "On the Stability of Network Distance Estimation", Proceeding of ACM SIGMETRICS Practical Aspects of Performance Analysis Workshop (PAPA 2002), also selected to be published in ACM SIGMETRICS Performance Evaluation Review (PER), September issue, 2002. [Tony McGregor]
  • Gave Jeff Olenchek from uwn access to the on-demand throughput tests; later in the period wrote him letting him know it was fixed after the amplet upgrade. [Tony McGregor]
  • Ian Graham from Waikato University and Dave Miller from Endace arrived from New Zealand and participated in the configuration and installation of the OC192 monitor (which has a Dag6 card).  
  • At the 10Gigabit Workshop, met with Ian Graham about OC192 development and he reviewed my slides for the 10Gigabit Workshop. I included an image of the new DAG6 card which will measure at 10GigE/OC192 speeds. [Ronn Ritke]     
  • E2E TAG call. The Pipefitters Project is interested in using AMP data for their E2E project. Matt Zekauskas 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. Matt let me know that he will ship the SC2002 Surveyor to SDSC - so SDSC will have one. [Ronn Ritke]
  • Met with Gabor, from Hungary, while he was visiting UCSD. They are interested in hosting an NLANR 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. [Ronn Ritke]
  • Called Dr. Xiao Chen at UCLA (statistician) about the current sampling research that Justin is working on. She responded that she is interested in working with us on research for a paper. [Ronn Ritke]
  • Tried again to contact Simone Triglia to inquire about her problems with data transfer.  No response yet; will keep trying.  [Cooper Nelson]
  • A delegation (of three) 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 1-2 AMP monitors and in developing some local measurement capabilities.  
    • Conducted a tour of SDSC and NLANR/MNA and met with the delegation; talks included a discussion of network systems.  An enjoyable experience for all and will likely result in a collaboration with that center as well as their joining the AMP mesh (possibly hosting two AMP monitors). [Ronn Ritke, Bud Hale, Jim Hale]
    • Met with the Taiwanese delegation (re NLANR/MNA and HPWREN), a meeting Peter Arzberger arranged. One of the better meetings, as they are working on issues very comparable to ours, for ecology settings in Taiwan, and were real contributors in the discussions.  [Hans-Werner Braun]
  • At the Fall I2 Member Meeting:
    • The NLANR Program Manager from NSF, Greg Monaco, was at the Fall I2 Member Meeting; we discussed current activities and some future plans.  Alan Blatecky (Program Manager at NSF) joined my conversation with Greg. Alan suggested doing some PR on the international collaboration model. The model consists of a country hosting an NLANR monitor and one or more groups from that country working to create a local measurement infrastructure and making that data publicly available. The local measurement infrastructure data adds to the data that can be used to detect and isolate network problems (see note on talks with group from Brazil). In time I want to write a paper that might include both the measurement infrastructure meshes, in addition to the "people meshes" that help make this possible. [Ronn Ritke]  
    • Alexandre Grojsgold from RNP in Brazil talked to me about hosting an NLANR AMP and possibly a PMA machine. He has had request for packet header trace data from researchers. An NLANR 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. That group of Graduate students can provide feedback on any modifications/additions to the existing AMP software.  [Ronn Ritke]     
    • I also spoke with Wendy Huntoon and Rick Summerhill about the possibility of NLANR AMPs at the GigaPOPs. The plan is to start with AMPs at 2-3 GigaPOPs and then proceed from there. [Ronn Ritke]  
    • Also talks with groups from Thailand, Taiwan, Brazil and Mexico. The CUDI project in Mexico has an OC3 connection to CalREN at SDSC. The CUDI NOC (located at a University in Mexico City) is interested in hosting an NLANR AMP monitor. The Uninet group from Thailand invited me to their December meeting, which may fit well with another project in Thailand that wants to leverage NLANR/MNA measurement work in that country. Followed up later with the CUDI NOC.  [Ronn Ritke]   
    • I spoke with Rich Carlson from Argonne National Lab. He is working on end-to-end diagnostic software and is interested in the possibility of using AMP data. [Ronn Ritke]
    • Eric Boyd from the I2 E2E Pipefitters project is also interested in using AMP data. Tony McGregor (our AMP manager) and I have given Eric's request for AMP data a positive response. We are waiting for Eric to respond with further details. [Ronn Ritke]
  • at SC2002
    • Met during SC2002 in Baltimore with Jim Ferguson and John Towns about possible joint MNA/DAST projects (joined by Greg Monaco).  [Ronn Ritke]
    • Networking researchers in Brazil may be interested in hosting both an AMP and a PMA machines.  [Ronn Ritke]
    • Sent an email to Manhee Lee asking how many hard copies he wants (and warning him that he's on the cover).  He wrote back that his colleagues/friends at SC2002 were quite pleased upon seeing the new international collaborations issue, especially with him on it.  (When Ronn received the copies, he also took some over to the KISTI booth at SC2002.)  Have received a lot of positive feedback about this issue (especially the much improved - thanks to Ben -layout and images).  [Maureen Curran]
  • Attended the Internet Measurement Workshop (IMW)in Marseilles, France. Great opportunity to meet with almost anyone we work with in the field, and so I did.  [Jörg Micheel]
  • Helped Cedric Westphal of Nokia IPRG (former Ipsilon) with access to trace data to compute distributions of IP address prefixes. Got him in contact with Christophe Diot of Sprint ATL's IPMON group, in the hope he can be helped there.  [Jörg Micheel]
  • Made another contact with a researcher at the University of Rostock (Baltic Sea, Germany) and they sounded very interested.  Whether they can make it with a paper is not clear, certainly they will be looking at attending.  [Jörg Micheel]
  • Met with Dr Hank Dardy and Basil Decina of Naval Research Labs in the Washington DC area. I got to have a look at all the interesting projects they are pursuing, including the new ViPR video conferencing system from Marconi, which is, at last, an excellent quality, easy-to-use system. Downside is that this lab is the very evangelist of ATM deployment, and, to my shock, is pursuing ATM even at OC48c and OC192c link rates.  Never mind, I also got to see some fabulous HDTV demonstration, which is convincing enough that there is some utility in moving that direction in technology. [Jörg Micheel]
  • Quite a bit of feedback from remote on PAM. Brynjar Viken came through with his draft on a paper asking for comments. Interestingly, the paper is in the same area as ours. We had some ideas about writing a joint paper some time back, it just never materialized. I think that ours and their contribution will be interesting to look at side by side.  [Jörg Micheel]
  • While in New York had meeting with Carter Bullard from QoSient, who is very active in the IETF IPFIX working group, and the author of a sophisticated commercial flow engine, freeware parts of which are available as Argus. [Jörg Micheel]
  • While in the bay area, gave Bruce Penrod of EndRun Technologies a call, learned that they are just about getting ready to have their Japanese version of the CDMA antenna tested, this would provide for another opportunity to get the measurements going that John Hicks started almost a year ago, across the TransPAC. [Jörg Micheel]
  • Spoke with Don Mitchell about current NLANR activities, positive response from the Russia trip and Bob Grossmans data mining application. [Ronn Ritke]
  • Planning with Peter Arzberger and Teri Simas for the upcoming PRAGMA Meeting; spoke with Man Hee and Kitamura about the meeting.   [Ronn Ritke]
  • Working on issues related to IBM disk reliability. Contacted the Center for Magnetic Recording Research (CMRR) on campus.  I learned that a major portion of the funding for the CMRR comes from the Sloan Foundation. That is an organization of major industrial players in the magnetic recording industry. Also some funding comes from US government agencies. CMRR conducts studies and research into many areas of magnetic recording. I discussed our concerns with a major researcher there: Gordon Hughes. CMRR has conducted major hard disk reliability studies, some in collaboration with IBM.  They are currently involved in some collaboration with IBM and Gordon is working with a IBM storage division vice president: Roger Hoyt. 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. Also Gordon is gathering some study material related to many of these issues that I will be able to pass on to interested people in NLANR/MNA (and HPWREN). As this collaboration developed, Gordon Hughes is working on getting failed disks analyzed for failure causes. He and I will be working with our supplier, RackSaver (formerly CPP), with the help and advise of the reliability people under Roger Hoyt of IBM, to get the analysis done. (This activity was brought about by the high number of disk failures that both AMP and HPWREN were experiencing.) [Bud Hale]
  • NLANR-coordinating conference call. [Hans-Werner Braun, Ronn Ritke, Mike Gannis]
    ~ ~ ~ PAM2003 - NLANR/MNA hosting

    Ronn Ritke, chair; Tony McGregor and Jörg Micheel, co-chairs

    The quality of submitted papers 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.  

  • Made some major layout/format changes to the PAM2003 Web pages. Made multiple modifications and added new content(including a tentative agenda)(approx. several times a week), regarding authors' instructions, deadlines, adding links, etc. Changed the notification line(s) in the PAM2003 navbar as needed to announce changes, deadlines, new content, etc. [Maureen Curran]
  • Numerous discussions with Nevil Brownlee (Univ. of Aukland, and CAIDA) re various aspects of PAM2003; liaised the content of these discussions to the "local" organizing committee (Ronn, Tony, Jörg). [Maureen Curran]
  • Keeping the Program Committee members updated regarding changes and additions to the Web pages (and deadlines); created email fanouts for distribution.  [Maureen Curran]
  • Created a flyer/PDF version of the CFP and Fedexed them to Jörg to distribute at IMW in Marseilles; also posted it on the Web pages.  In an effort to encourage student participation -  distributed the CFP on campus to appropriate departments (graduate coordinators and undergraduate advisors), CAL-(IT)2, and sdscusers (many of whom each teach, take classes, etc. in the CSE department).  Also sent it to itc@i-teletraffic.org and ippm@advanced.org and to the Program Committee for fanout. [Maureen Curran]
  • Got the EDAS site  (software system for conference paper submission) setup for PAM2003, after collecting the pertinent information about conference from Ronn and Maureen.  [Cooper Nelson]
  • Handled and resolved many issues with EDAS (the paper submission system). <Great software program, hideously underdocumented, causing much time and trouble.>  Wrote additional instructions on using the EDAS system to facilitate use (was quite problematic); worked extensively with Henning Schulzrinne (Columbia, EDAS) on several aspects of the system. The paper submission deadline had to be extended due to the numerous submittal problems. The situation was referred to (by EDAS) as "the discovery by random chance problem"; apparently many things have been corrected and documented now.    [Maureen Curran, Cooper Nelson, Jörg Micheel]
  • Coordinating the paper review for PAM2003. Received close to 100 submissions thus limiting the number of reviews per paper to two. Developed the review questions with the committee. There was a very interesting discussion on the scoring of the papers and we have arrived at what I believe is a good way to highlight the 20 papers we really want (and can afford to accept, given the timelines) while cutting down on the time to be spend on sorting the borderline cases. We want to provide quality feedback to the submitters.  [Jörg Micheel]
  • Reading, reviewing, and commenting on the submitted PAM abstracts (10-11 papers assigned to each). [Tony McGregor, Jörg Micheel, Ronn Ritke]
  • Planning:  met with Donna Turner about food planning for the PAM2003 Monday night dinner. We have the location at UCSD picked. Donna will make sure that area is not watered a week before and arrange to get tents for the food.  Nancy Jensen has helped with the PAM2003 beach towels. They have arrived and now are in my office for safe keeping. The beach towel will be a giveaway and we may add a frisbee. Worked with Gail Bamber and Nancy Jensen on the PAM2003 workshop beach towel.  [Ronn Ritke]
  • Began preliminary development of the registration process.  Multiple meetings included meeting with Donna Turner and in particular with Cindy Lee re SDSC's new registration audit trail system. Cindy Lee, the computer person for the SDSC business office, outlined the set-up and processing of registration for PAM.  SDSC is setting up a conference/meeting registration system which we are going to be a part of (actually, we be the first "users").  All registration info and payments will be done on-line, including the capability to do UC (any UC) recharges. The system was presented to us as being in a "up and ready to go" state (which we found out later was (definitely) not true). Cooper and Chris began the back-end work. [Maureen Curran, Cooper Nelson, Chris Gross, Ronn Ritke]
  • Setup account and tools to create registration page for PAM2003.  [Cooper Nelson]

    ~ ~ ~ Students
  • Helped the new students <Chris and Ben> with setting up various things/answering various questions related to FreeBSD.  [Bill Gahr]
  • Met with Justin weekly re his new sampling graphs and progress on the abstract for PAM submission. [Maureen Curran]
  • Gave a presentation at the ANR students presentations on my sampling work.  [Justin Fields]
  • Helping out the new students <Chris and Ben>. [Justin Fields]
  • Discussions with Todd about the various student projects; his help has been invaluable (to both Ben and Chris). [Maureen Curran]
  • Gave Ben and Chris a CGI-perl/gnuplot presentation.  [Todd Hansen]
  • Assisted Chris with questions he had regarding security, operating system selection, and unix lore.  Also discussed flow metrics with him. [David Cheney]
  • Met with Chris a couple times per week regarding his PHP tutorial, the PAM registration pages. and other activities.  Since Jörg had suggested he look at the CAIDA pages, took the opportunity to introduce him to Nevil, who recommended that he read the metrics tutorial that he (Nevil) had done a couple of years ago. [Maureen Curran]
  • Met with Ben Reesman and Maureen to go over Cichlid plans. Ben seems very enthusiastic about the future of Cichlid. [Ronn Ritke]
  • Met with Ben a couple of times per week re his work with/on Cichlid; introduced him to Brad Huffaker (Todd's suggestion) and contacted Mike Bailey. Worked with Ben on a quick rectangular shape PAM2003 logo/wording for the list page on EDAS (forwarded to Henning). [Maureen Curran]
  • Introduced Ben to Jeff Brown (original developer of Cichlid, arranged a first meeting), to kc claffy, Young Hyun (new super viz person), and Brad Huffaker of CAIDA and began to arrange a meeting with visualization folks at SDSC.  [Maureen Curran]
  • In preparation for working with Ben on his references for the presentation, did a search for complete reference/citation formats (one of Ben's is a dissertation) for CS and found a great resource, which covers all cases and has examples for each (by IEEE Canada), forwarded it to all, especially the students. [Maureen Curran]
  • Met with Ben (along with Todd) several times re the direction of his study and have him explain his findings.  Also discussed the future direction of study and worked on his presentation.  [Maureen Curran]
  • Learned to use Adobe InDesign in the process of working on the NATimes.  [Ben Reesman]
  • Learned a lot from Bill Gahr (HPWREN student) about concise descriptions of trees and different data representations and algorithms.  [Ben Reesman]
  • Met with Bill Gahr to discuss research possibilities for his planned time at a Japanese university. One idea is to study the telescience application that runs between Osaka Japan and SDSC. Sent him contact information for Kitamura, who is the APAN NOC Measurement Chair, and went over the locations of relevant AMP and PMA sites that may help in that research. [Ronn Ritke, Todd Hansen (by phone)]

  • Other Activities

  • Tony and Matthew travelled to SDSC for a week in early December.  Lots of meetings with various people here. Basically catching up on where people are at and tidying up loose ends. [Tony McGregor, et al.]
  • Met while Tony was in town to strategize on some NLANR/MNA next steps. [Hans-Werner Braun, Ronn Ritke, Tony McGregor]
  • Began an inventory and storage of equipment in currently located in the office/work area (AMP, PMA, and HPWREN equipments/tools). We need to make serious storage arrangements. [Jim Hale]
  • Liaised with NSF regarding Tony and AMP repackaging. To follow up at the NSF PI Meeting Jan 9th and 10th. [Ronn Ritke]
  • Jim Hale became a full-time UCSD/SDSC employee (50% NLANR/MNA, 50% HPWREN), as the AMP/PMA assistant systems administrator.  

    ~ ~ ~ Staffing and Admin:
  • Things are continuing to progress on getting more of my time available to the AMP project.  I've been involved with discussions at the Waikato end. [Tony McGregor]
  • Meetings and talks concerning the Waikato subcontract and PAM2003 planning. [Tony McGregor, Ronn Ritke, Mary Dimeglio]
  • For the PAII 100% position, arranged and conducted interviews with the top three applicants; htmled the results (using the posting template that I designed), posted for review, participated in follow-on discussion.  Final selection has been made. [Maureen Curran]
  • After interviews and decision (PAII), we completed the selection process paperwork for HR and discussed the job description for Dave's PT work.  Rewrote Dave's job description, broke it down on the HR job card (overview, functions/tasks, knowledge/skills needed), sent completed form to Ronn. [Maureen Curran, Ronn Ritke]
  • Did some follow-on work regarding the student opening(s). Emailed notification to the students who did not make the final cut in the selection process. Sent about 20 of the student resumes/CVs to Eric Terril/Lisa Lelli of SIO for consideration in their search for a student to do some coding, etc. (sent a notification to the students regarding this). [Maureen Curran]
  • Discussions with Jörg about some local student help with PMA analysis. [Jörg Micheel]
  • Met Ben Reesman one of our new student employees. Introduced him to Sandy Davey, Amy Han and Rachel Chrisman so he could obtain his office door key, SDSC door key and SDSC accounts. [Ronn Ritke]
  • Miscellaneous HR tasks: arranged for Maureen to handle Dave's time sheet and to be able to sign for Dave, Justin, Ben, and Chris. Spoke with Barbara Carstens about some HR paperwork and job descriptions. Followed up on some possible student support for Jörg. Spoke with Gail Bamber, Mike Gannis and Joe Czech about the GA conversions to UCSD. Booked airline tickets for Jim to attend AMPATH. Met with Sharon Kelly about some HR paperwork (loads of HR paperwork), and about job postings and current status of recent and planned new hires for the NLANR project. [Ronn Ritke]
  • Payment authorizations, MNA budget issues, REU payment authorizations, purchase orders, meetings with Ronn, travel vouchers, account reconciliations, summary reports, Waikato funding/subcontract, and other budgetary matters.  [Mary DiMeglio]

  • Inclusive list of sysadmin tasks

  • The following systems administration tasks apply to both the PMA and AMP projects.  They are performed primarily by Bud Hale, with the assistance of Jim Hale.
  • 1. Maintain an on-going and continuous rapport with all NLANR/MNA remote site people. Contact site people on a regular basis and inquire as to the activities at each site. This is necessary to keep up with personnel changes and network configuration changes at each and every site.
  • 2. Provide educational and informational materials as need by remote site personnel to garner the motivation for support of the equipment on site and the needed connectivity.
  • 3. Maintain the remote site database and enter any and all changes to equipment, connections and contact personnel.
  • 4. Maintain records of the equipment deployed as well as the NLANR/MNA equipment at SDSC.
  • 5. Maintain an on-going relationship with vendors of equipment and parts needed to support the entire NLANR/MNA infrastructure.
  • 6. Continuously monitor the performance of all NLANR/MNA equipment on a 7 day, 24 hour basis. Perform test and analysis to anticipate equipment failure. Monitor equipment to determine the fill rate of data storage.
  • 7. Maintain a rapport with SDSC personnel in all departments to garner support needed for NLANR/MNA equipment installation, NLANR/MNA network connections, work stations and spaces.
  • 8. Continuously monitor the operational status of SDSC equipment such as the HPSS to determine the usability by NLANR/MNA functions.
  • 9. Continuously monitor the accumulation of NLANR/MNA data and perform NLANR/MNA data archival as needed.
  • 10. Maintain a supply of spare parts and/or maintain sources of spare parts supply.
  • 11. Perform on-going equipment repair on a 7 day, 24 hour basis as needed.
  • 12. Run the procedures to accomplish remote site system updates on a regular basis.
  • 13. Monitor NLANR/MNA server operation and performance to anticipate equipment failure and/or software performance problems.
  • 14. Maintain an awareness of industry progress to anticipate obsolescence. Perform tests of new components as necessary to determine operability with NLANR/MNA systems.
  • 15. Analyze connection anomalies to develop corrections as needed. Maintain records of corrections applied to aid in future problem resolution.  
Back to the Top