Maggie-NS Tasklist

 

March 4, 2005

 

NEXT MEETING: March 11, 2005

Awaits something, also provides start of wait date

{Done or Drop. Is deleted 2 months after task finished or dropped}

– Person doing the work

Bold face - task in progress, or for discussion at meeting

Highlights are recent developments

 

Action Items   

  1. PingER –SLAC contact: Jerrod

    1. Provide tools to analyze and create plots of PingER data - Jerrod, Usman Ahmad Mian [ahmed.usman@niit.edu.pk]

      1. Tie into webservices/SQL access to data to make a much more flexible distributed way of selecting analyzing and displaying data

      2. Regional plots for executives - Usman Ahmad Mian, Fawad

        1. Get plot for single region {Done 12/804} - Usman
        2. Restore data {Done 1/27/05} - Les, Connie, Jerrod
        3. Place a link in the pingtable to the plots created by Usman's script. { Done Jerrod 1/27/05}
        4. Integrate with Pingtable.pl {Done 2/15/05}
        5. Improve labeling
        6. Fix bug that labels axis as RTT for all metrics (Usman has a fix 3/4/05)
        7. Add percentiles (Fawad has defined, and Usman is working on 3/4/05)
        8. Add log scale option {Done 3/1/05}
        9. Add fits (Fawad has defined 2/4/05)
        10. Provide access to data {Done 2/28/05}
        11. Pass parameters from pingtable.pl to PingerHistoricalData.htm (Defined need, Usman will look at 3/4/05)
        12. Extend to multiple regions like Figs 2, 6, 8 in ICFA report
      3. Fix bug that causes Oct 2000 to be displayed twice in allmonths (Ashar will work arounJ, Jerrod has looked at but does not see the cause 9/23/04, still there 11/24/04, Jerrod will investigate, Les checked the problem is not in pingtable.pl but in the data it is presented with 11/27/04) {Fixed by Les 12/14/04} - Jerrod
        1. Re-run data {Done 12/14/04} - Les
    2. Develop a package to make PingER installs of “monitoring sites” a simpler process.
      1. Set up a website for PingER (would need very close collaboration with SLAC, possibly once a student has arrived at SLAC)
      2. Extend Pinger Gathering to make it hourly so data is more “real-time” (in progresss)  - Nazir, Jiri
        1. Plans are for NIIT to download PingER historical data once a week (or so) and test their tools against downloaded data.  This practice mirrors SLAC running the tools in-house similar to how the tools will be used once in production.  
    3. Convert offsite.nodes to a database format
    4. Pingtable.pl
      1. Review PingER2 and see how to incorporate
        1. Request & get tar ball from Warren (virus checker removes tar ball from mail, 5/16/04, got via anon FTP) - Les
        2. Install and check out - Warren
      2. Provide machine readable access to monitoring host participants etc. - Maxim
      3. Make pingtable.pl available by FTP (Raja is testing 8/3/04) {Get data from web instead - 9/22/04} - Jerrod, Les
    5. Web services access to Guthrie - Jerrod
      1. Fawad reports there are still duplicates 10/20/04 - Jerrod
      2. Missing data in locations.txt (Fawad will work around, Jerrod will see if Fawad can update the Guthrie database himself to add the missing information 9/23/04) {Done 11/24/04} - Fawad, Jerrod
      3. Fawad needs clarification of node numbering (11-24-04) {Done 12/7/04} - Jerrod, Fawad
      4. Provide web services access to host descriptions - Fawad
    6. Get more hosts monitored from Pakistan, work with Broadband Networks (got hosts for all pk regions, also got developing regions hosts - 3/4/05) - Fawad, Les
    7. Triangulate position of an unknown host
      1. Define project, do literature survey - Fawad, Les
  2. NDT –SLAC contact: Connie
    1. Next steps will be to use it following detection of an "event" to gather more information on the problem

  3. NETRAT –SLAC contact: Connie
    1. Work on change detection and alerting of changes in the network performance (good idea for a summer project then carry back to NIIT)
      1. Evaluate the PCA multivariate method of Crovella - (9/23/04 Waqar/Azam/Arshad Ali)
        1. Get algorithm working for two variables {Done 12/7/04}
        2. Extend to five ABwE variables - Waqar, Azam
        3. Try it out with SLAC data (plan to have by Dec 8th) - Waqar, Hamad
          1. Send SLAC data {Done 12/8/04} - Les
        4. Compare with alternative techniques
        5. Extend it to multiple routes - Waqar, Hamad
        6. Develop visualization tools
        7. Get meaningful results from the anomaly detection and provide results and possible plots by Monday (12/16/04)
        8. Add another parameter, multi-route analysis, to this testing (1/6/05)
        9. Azam is having problems are fetching the abwe data {Done 2/25/05} - Azam, Connie
        10. Interpolate and trim data (Les send information on Maxim's wavelets 3/4/05) - Azam
        11. Adjust current matrices used with this problem detection method attempting to detect the missing anomalies. (1/27/05)
        12. Implement new problem detection matrices  to (1/27/05)

      2. Implement H-W technique - Felipe
        1. Provide interpolation script and save data - Felipe, Connie
        2. Estimate parameters - Felipe
        3. Get HW to work & compare with Brutlag - Felipe Mahesh
      3. Evaluate Burgess method - Mahesh
        1. Put forward proposal {Done 11/24/04}- Les, Alf
        2. Take SLAC IEPM/ABwE data, massage (interpolate) to provide measurements ar regular intervals, and format for CFE - Need person at NIIT?
        4.
    2. Work on end-host and network monitoring in particular with web-100 - Aziz Allaudin
      1. Decide what is useful, possible to monitor - Aziz
      2. Reduce reliance on web-100 - Aziz
      3. Figure out what can be measured/recorded and come up with script to measure and record (Les sent suggestions for host system 1/27/05) - Aziz
        1. Look at the "slow" tool (need to remove sending email to SLAC) - Aziz
        4.
      4. NIC Monitoring - Aziz
        1. Find a way to retrieve NIC parameters for both Windows and Linux systems.  (12/16/04)- Aziz
        2. Developing the Linux version of NIC monitoring.  Sent Les a version of the Linux Nic monitoring tool to test <needs root privs. to execute>(1/6/05)
        3. Develop the same monitoring tool  for Windows systems(1/6/05)
      5. Create web page describing problem and providing access to relevant current information (work in progress 1/27/05)- Aziz
    3. Add to the detection of network problems the gathering of relevant info. For the complete path.  The idea is that along with the alerts, we provide a bit of diagnostics to present to the sites seeing problems so that we don’t just send messages reporting the problem but provide info that could help them solve the problem as well. - Fahad Khalid
      1.  Filter alerts before reporting
  4. IEPM-LITE –SLAC contact: Connie
    1. Re-architect IEPM-LITE to make it more portable  - Connie (5/26/04 - in progress.)
      1. Once the redesign is completed, Connie needs to have a few cgi scripts written to facilitate adding nodes to the IEPM-LITE tests.
    2. Complete a distribution kit for Iepm-Lite.  (Jerrod)
    3. Generate ABWE's config. file from bandsrc/slaconly/aliases.  Include with this the ability to automatically detect/reload a more current version of the config. file.  (Jiri)
    4. Package ABWE for distribution. NEEDED: a way to run ABWE every 5mins using gnuplot.
    5. Send NIIT the structure for building an Oracle database to handle the IEPM-LITE data {Done}
    6. Install IEPM vsn 3 Mati-ur-Rehman
      1. Send Mati-ur-Rehman Khan IEPM install package (11/28/04) {Done 12/8/04} - Connie
      2. Mati will try and install (having problems with mysql pre-requisite 1/4/05, can Jerrod assist 2/15/05) {Done 2/18/05} - Mati-Ur-Rehman
      3. Load data for 4 sites being monitored from NIIT (Mati understands how to do it 3/4/05) - Mati
    7. Improve traceanal reporting for fluttering NIIT traceroute - Fawad Nazir, Les, Connie
      1. Fix lack of name service for NIIT nodes {Done 12/4/04} - Arshad, Les, Azher Amin
      2. Understand fluttering routes (email sent to SingTel 12/4/04) - Les
      3. Describe problem and possible approach for NIIT {Done 12/13/04} - Les
      4. Why 1st 3 octets or AS don't eliminate route change (walked through the code indentified where things go wrong 3/4/05) - Fawad, Connie
      5. Traceroutes from NIIT to SLAC differ from the SLAC to NIIT route.  Les will forward Fawad the URL to the online traceroute utility used at SLAC.  This allows Fawad to perform the reverse traceroute from SLAC to NIIT to study this problem. (12/16/04) {Done 12/17/08} - Les
    8. Extend Plot Lat/Long to IEPM data - Sufyan
      1. Jerrod will specify - Jerrod, Les {Done specified 1/27/05} - Jerrod
      2. Implement using the 'readonly' account on iepm-resp- Fawad
    9. Add reverse abing to IEPM - Connie
      1. Collect data - Connie
      2. Document how to access - Connie
      3. Provide forward and reverse graphs (awaits collecting data) - Connie
    10. Get abwe reflector running reliably at SLAC on iepm-resp (it is very reliable, so not high priority, awaits reverse abing 12/14/04) - Connie
  5. AMP project
    1. Contact AMP/NLANR folks to see if they are interested {Done 3/4/05} - Les, Fawad
    2. Get IP information on host to AMP folks - Fawad
    3. Communications etc.
      1. Improve reliability of monitor.niit.edu.pk
        1. Monitor and record availability (agreed to 11/24/04) - Arshad/Fawad
      2. Travel
        1. Fawad's documentation has gone to Washington (12/8/03 now wait 3 weeks)
          1. Applied for visa 2 weeks ago, no update.  He expects to have it back by the end of December/ 1st week in January.(12/16/04) {Done 2/10/05}
          2. Leave Pakistan for SLAC 2/19/05 {Done 2/19/05} - Fawad
        2. Dr. Arshad looks to visit SLAC starting April 4/5 for 10 days.
          1. Get invite from SLAC {Done 12/14/04} - Les
          2. Awaits visa {Done has visa 3/5/05} - Arshad
        3. Jerrod Travel Dates Feb. 6-20 2005 (12/8/04)
          1. Create list of tasks/projects for Jerrod's visit-FAWAD (done 1/17/05)
            1. Series of presentations on overall perspectives, objectives for IEPM/MAGGIE  project
            2. Meet with ISPs to understand current Pakistan communications networks and plans and make contacts for future
            3. Meet with various NIIT research groups to understand and learn about projects
            4. Create draft {Done 12/12/04} - Fawad
            5. Send back feedback (12/16/04)- Jerrod (done 1/27/05)
        4. Les decide on dates, submit DOE travel request - Les
          1. Visa request sent 2/18/05
          2. Seats reserved 2/19/05
        5. Aziz
          1. Verify green card will still work (2/10/05 I131 awaits reply in 4-6 weeks from US State Department) - Les, Aziz, Robyn Mosset
          2. Get invitation letter to Azia (Les sent draft to Robin 3/4/05) - Les
      3. Recover IEPM disk files - Jon (done 1/20/05)
        1. The PingER data that was lost on this disk has been retrieved and was able to be stored using the same directory structure as we have before the disk crashed.  The pingtable has been recreated and is now functional.
    4. Next proposal
      1. Letter for Dr Atta-ur-Rahman - Les, Prof Arshad  {done 1/5/05} - Arshad, Les
      2. Verify that the letter was mailed to Dr. Atta-ur-Rahman  1/27/05, need phone and street addresses (not POB) for FedEx) {Done 2/10/05} - Les
      3. Review work breakdown structure and provide feedback {Done 2/10/05} - Les
      4. Look at proposal for a testbed funding from NTC - Arshad
    5. Jiris Nearest Abroad Neighbor (see http://202.83.166.170:8080/examples/servlet/NAN) Fawad, Jiri
      1. Specify what the result might look like for a couple of sites - Jiri
      2. Provide table for each monitoring host, for each remote host, by TLD, add extra metrics (min, avg, std, distance (identify distance as great-circle route) etc.) - Fawad
      3. Provide a way to read PingER data from monitoring sites in near-real time - Jerrod, NIIT
      4. Add traceroute measurements to PingER monitoring sites - Jerrod
      5. Provide country information on hosts - Les, Jerrod
      6. Plot received data onto graphs, see http://202.83.166.170:8080/NAN.html  {Done 2/11/05} - Fawad. 
      7. Review {done 2/11/05}  - Les, Fawad
      8. Trim the number of decimal points
      9. Make legend visible always
      10. Add extra sites - Jerrod, Fawad
      11. Port to SLAC