Attendees
Anjum, Kashif, Raja, Les. Joun and Umar were not present/
General
Arshad has put together a letter from the NUST Rector to the VC of UM (or is it UTM) in Kuala Lumpur concerning a collaboration of Pakistan, Malaysia and SLAC.
UNIMAS is starting to monitor other Malaysian hosts outside the regular Beacons. There is a problem with adding the non Beacon hosts. The pinger.xml file seems to keep getting replaced with an old copy.
UNIMAS are looking at whether they can fund the NUST people for a workshop in December at UNIMAS in Sarawak.
From Arshad: Rector has agreed to support one person visit to SLAC. Pl select the best candidate. Anjum will take the lead on this.
Deploy another node in Faisalbad University. Have a POP node, this will give us more information. This has been installed and sent to Faisalabad. They plan to have it up in a few days.
HEC Report - Anjum, Amber and Imdad
Kashif is taking this over. Better to prepare from 1-30th of month, then submit on 5th of following month. It has been completed for September and sent to HEC and the SEECS email list.
Pakistani Hosts
LSE is unwilling to install PingER due to security concerns.
FSBD POP is working well. We believe they must have replaced the backhaul network with their own network with no commercial traffic on it.
Kashif looked at FSBD pingtable.pl data for Sep 4th and the RTT from FSDB to PWR as "too" stable to be true. It is working fine now but the large values seen just on Sep 4th are distorting the overall averages. This appears to be since only 2 measurements were made that day.
Pakistani monitors: good=35, no_data=9, partial_data=3, problems=12, total=47. Pingerpwr has an IP address issue (they need to change their address). HEC is working on it. PIEAS and IIU have new installations in progress. SAU is in the midst of a strike.
RTT values from HEC-Lahore to HEC-Quetta are very high since 15th September 2012, i.e. went from ~35ms to ~150ms To analyze this anomaly Kashif ran the traceroute from HEC-Lahore to HEC-Quetta and also compared it with the traceroute from HEC-Lahore to HEC-Karachi. Up to 05 hops the route is same and afterward the routes are entirely different. At hop number 05 Transworld Pakistan route the packet to an IP in France (Values got from: http://whatismyipaddress.com) and then packets come back to Pakistan at hop number 10 which is the reason for the higher RTT. Whereas in the comparative traceroute from HEC-Lahore to HEC-Karachi the values are fine. Therefore this issue need to be investigate at HEC-Islamabad level that why Transworld is routing the Quetta packets to outside pakistan resulting in higher RTT. This issue is same for HEC-Lahore to BUITMS-Quetta which implies every packet from HEC-Lahore to Quetta Region is going outside pakistan. Slides 1 & 2 of the attached spreadsheet (note that when there is 100% packet loss there is no successful measurement of RTT) show the losses and RTT from 4 Sep to today. It is seen the time series that the extended times started appearing Sep 8th - 10th and then turned on solid on Sep 15. There is evidence that it dropped back sometimes on Sep 30 and again Nov 3 and Nov 4th. It again looks to drop back on Sep 09 and 10.
Kashif has sent a report to Anjum who will send it to HEC and request a meeting early next week to address the issue.
It would have assisted to have had a log of traceroutes. Anjum will contact Zafar who was working on this. This was completed for SLAC but we need it for SEECS PERN POP hosts.
Non Pakistani Hosts
Sri Lanka is OK again the node namunu will be replaced by hunnas (it is running freeBSD and needs https). We are currently gathering data from both. SDSC is down they are rebuilding their PingER server. pinger.cdacmumbai.in is down for about 5 days. Mail has been sent to sitka.triumf.ca, there is no response. The UERJ host lost its hard disk, it is being replaced.
TULIP
Raja understands the CGI code in Matlab.. He is working on the MatLab version to add a lookup table to allow the choice of Alpha to be based on region and min RTT. When he has this working he will work on porting from MatLab to Mathematica.
It may also make sense to look at using a modified outer circle. I.e. instead of distance (radius in km)= 100km/ms*min_RTT(ms), use distances(radius in km)=alpha_max*100km/mx*min_RTT(ms) where alpha_max is the maximum alpha for landmarks to target in a given region.
Statistica analysis of Alpha
Raja agreed to provide a written write up of what he found out by analyzing alpha for various regions.
Future meeting time - Les
Next meeting: Wednesday 17th October 8pm Pacific time, Thursday 18th October 8am Pakistan time. Big topics will be deciding who comes to SLAC and the results of the email to HEC concerning the routing between Lahore and Quetta.
Following this I will be in Kenya starting October 19th thru November 3rd. So I propose the next meeting to be Wednesday November 7th at 8:00pm. Pacific time and Thursday November 8th at 9:00am Pakistani time. Note the change of time in Pakistan due to the end of daylight time in the US .on Sunday November 4th.