Netflow cc 20060228

From GEANT2-JRA1 Wiki

JRA1 Flow Monitoring - conference call agenda and minutes

1 Info
2 Agenda
3 Minutes
4 Conference call details

[edit]

Info

Date/time: 28th February 2006, 14:00 CET

Expected time of conference call: 60 minutes

[edit]

Agenda

netflow boxes update (check list of netflow boxes)
Anonymization task (reference: Netflow_anonymization)
perfSONAR architecture and netflow measurements correlation
netflow exchange format
AOB

[edit]

Minutes

Attendees: Hans Trompert (HT), Luuk Oostenbrink (LO), Alessandro Inzerilli (AI), Igor Velimirovic (IV)

netflow boxes update: all boxes (ZG, AM, MI) up and running. Short reports:
- Zagreb: NERD installed, can be accessed via http://193.198.229.186:8090/. For now it collect netflow export from 1 test router, with low amount of traffic.
- Amsterdam: IPFLow installed from binary file (no source available). There is a problem with AVICI routers, it seams that there is no SNMP communication between router and collector. Also, IPFlow can not re-send (replay) netflow export. There is a problem that IPFlow developers are not responsive on e-mail.
- Milan: flow-tools installed and running. Also, using flow-fanout netflow data exported to nfsen/nfdump tool which is also installed on the test box. There was however problem with NERD installation.
Anonymization task: we should use some of the existing methods/tools, listed on Netflow_anonymization. Also, anonymization should be one of the 'must have' features in tool evaluation.
perfSONAR architecture and netflow measurements correlation: there was a discussion by HT started on e-mail (see thread 'Netflow exchange format'). Regarding presented options we agreed that preferred one is a first one (exporting of row netflow data to the user) but it is the option that does not fit into perfSONAR overall architecture at all. Therefore discussion with perSONAR topology people should be started as soon as possible. Alternative is option 3, and the option 2 should be avoided. Also, referring to the CARNet experience with storing netflow data in a SQL database, there is a proposal to investigate such a solution as well.

Actions

AI to send to IV installation error from NERD (since CARNet installed NERD successfully)
IV to once again review the evaluation report ( Netflow tools evaluation guidelines v0.1), by the end of this week decide weather each partner should fill the report.
all should continue tool testing using netflow data from routers within their networks
AI to add paragraph about nfsen/nfdum anonymization capabilities on Netflow_anonymization page
IV to report on a solution of storing netflow data in a SQL database
all to participate discussion started by HT (tread 'Netflow exchange format') in order to prepare input for prototype team as soon as possible

Next conference call in two weeks time (13-17.02.2006.), to be announced next week

[edit]