Agenda for SEECS/SLAC meeting 8/21/2010

Pakistani case study – Fahad, Anjum

  1. PERN still down (unable to gather any data) is there any progress?
    1. no progress, person who was going to look at had a family emergency, hope to have next week
  2. Unable to gather data from uettaxila for last week
  3. PERN's request for progress report
  1. We have a working version of table.pl, it provides much info
    1. Some monitoring sites do not have a complete set of hosts, who is working on this
    2. NCP and IIU have a circuitous route
    3. PERN looks best as expected
    4. jitter is low
    5. unreachability is high
  • Need to get the beacons, Umar explained, Anjum has requested nodes to send xml and beacons file so can see if they are OK. The Pakistani monitoring nodes are supposed to get the beacon list from SEECS – Umar

Kashif Sataar of SEECS is looking at whether data is gatherable from the Pakistani nodes.

ping_data.pl - Fahad

ping_data.pl: extending diagnostics  - Fahad

  1. Display permissions in octal - Les
  2. Check validity of permissions - Les
  3. Check for free space - Les
  4. Check for libraries - Fahad

TULIP

Added perfSONAR landmarks (are they showing up on map) - Zafar

  1. 41 nodes out of 80 added. Others not added since could not find the URLs, will recheck in a few days.
  2. Apollonius and trilateration can fail for “inscribed” landmarks. Should add ability to ignore landmarks which are within a specified “short” distance of one that has already been used. Mark such landmarks as “dups”.-Zafar
  3. Qasim set it up so two or more landmarks from the same landmark should not be used. This programmed into trilateration algorithm. For Apollonius if do not get a result then change the set of landmarks.

TULIP GUI - Faisal

Working on the summary:
  1. It needs a refactorization.
  2. Added ability to hide landmarks.
  3. Apollonius integrated with circles & markers.
  4. Added Ajax based information box currently showing failures, anything else?
  5. Added distance into mouse click info window.
  6. Can we use to understand Apollonius failures?
Live GUI:
  1. Needs refactorization.
  2. Needs re-write
  3. Adding other geolocation estimates to GeoIPTools (GeoTools, IPLocationTools, whois ...) to come.
  4. View via Information box,
  5. provide check marks to select which extra geolocation estimates
TULIP map: - possible project
  1. Add disabled landmarks, Faisal is adding this
  2. Add traceroute & ping to my host and to selected host - Faisal will add.

Why does Apollonius fail - Farrah

Apollonius fails in 55 out 171 cases why?

  1. She is working with another incubator company in NUST so is excited about this
  2. Umar will talk to at length

MatLab - Zafar

Can't convert MatLab code to Java in an automated fashion.

  1. Zafar tried a tool to do this. It failed.
  2. Alternate solutions, rewrite the selected MatLab code, or use a MatLab server
  3. Looking at feasibility,
  4. Might ssh to a Linux Matlab machine and not implement the MatLab code again
  5. Getting Linux license - Zafar

Paper - Umar

Google waves alternate?
SVN for paper
Where do we go from here, do we await INFOCOM comments?

  1. Do we rewrite with a new story – Umar is looking at
  2. How much time can Umar spend on it?
  3. Fall semester starts Monday so Umar cannot drive.
  4. Need a person to drive the paper, in particular the first few pages with the storyline
  5. Zafar is volunteering to re-write the story line
  6. 1st step put together a story line in Google wave

PingER metric motion charts - Zafar

CSV files are updated, next is to integrate into motion charts.
A new index is needed for the IDI index.

  1. IDI replaces DOI from 2007 onwards.
  2. Hopes to have it done this week

Future - Les

Funded for $100K for each of 2 years for Yee and me to work on making perfSONAR measurements more easily accessible, easily navigable etc, plus access to end site data. We would like to free up some of SEECS resources (Faisal or Zafar) to get involved with this research.

  1. This is good news.
  2. Zafar returns to Pakistan at the end October, then maybe return early January.
  3. Zafar will apply for entrance to Stanford in 2011.
  4. Zafar will work as an RA.
  5. There is a possibility of extending. He would need to return first.
  6. Can the funding from perfSONAR be used for the student.
  7. Meanwhile Anjum will take the lead on funding from NUST side.

Project Ideas

  1. Tulip map update with ping & traceroute
  2. TULIP database GUI, need to understand phpmyadmin, Umar's php and integration at SLAC
    1. There are problems with this approach, in that php is not accepted by security at SLAC and phpmyadmin is not supported at SLAC
    2. We will use mysql workbench client.  Failsal has competed this apart from writing to the database
  3. Identify duplicate landmarks (a landmark within some selected distance of landmarks already looked at) in the TULIP database and mark them as duplicates. Once this is done modify reflector to provide an option to ignore duplicates.
  4. Modify table.pl to enable use of table.pl for displaying a full matrix of Pakistani monitoring hosts.

Future meeting time –

Thursday 10am?

  1. Umar is strongly in agreement.
  2. Anjum can make it.
  3. Anjum contact Fahad.
  • No labels