Project

General

Profile

Actions

Feature #251

open

batadv-vis: Add support for B.A.T.M.A.N. V throughput

Added by Russell Senior almost 8 years ago. Updated over 7 years ago.

Status:
New
Priority:
Normal
Target version:
-
Start date:
05/14/2016
Due date:
% Done:

0%

Estimated time:
40.00 h

Description

I'm experimenting with BATMAN_V on the lede-project revision reboot-231-gf8abb68 with batman-adv and alfred. The batadv-vis program reports:

root@mesh-test1:/# batadv-vis -v
batadv-vis 2016.1
VIS alfred client

With a three node test network, mesh-test1 and mesh-test2 are linked via both ethernet and wifi ibss mode, mesh-test3 is linked only with wifi, and I get odd looking results from batadv-vis:

root@mesh-test1:/# batadv-vis | grep -v TT
digraph {
subgraph "cluster_00:0f:b5:97:28:9d" {
"00:0f:b5:97:28:9d"
"00:0f:b5:0c:e0:84" [peripheries=2]
}
"00:0f:b5:97:28:9d" -> "00:0f:b5:0e:71:5b" [label="2.550"]
"00:0f:b5:0c:e0:84" -> "00:12:cf:83:7b:09" [label="6.711"]
subgraph "cluster_00:0f:b5:0e:5d:8f" {
"00:0f:b5:0e:5d:8f"
"00:0f:b5:0e:71:5b" [peripheries=2]
}
"00:0f:b5:0e:5d:8f" -> "00:12:cf:83:7b:09" [label="6.711"]
"00:0f:b5:0e:71:5b" -> "00:0f:b5:97:28:9d" [label="2.550"]
}

The numbers don't seem to ever change, and are way higher than what I would expect from ETX. I'm informed, not surprisingly, that BATMAN_V doesn't use ETX. Whatever metric is used, it might be nice to have it reported.

Actions #1

Updated by Sven Eckelmann almost 8 years ago

  • Tracker changed from Bug to Feature
  • Subject changed from batadv-vis reporting odd looking link metrics with BATMAN_V to batadv-vis: Report used metric to differentiate between throughput/TQ
  • Assignee set to Simon Wunderlich
Actions #2

Updated by Sven Eckelmann almost 8 years ago

  • Subject changed from batadv-vis: Report used metric to differentiate between throughput/TQ to batadv-vis: Add support for B.A.T.M.A.N. V throughput
Actions #3

Updated by Antonio Quartulli almost 8 years ago

IMHO when using B.A.T.M.A.N. V we should just report the real metric (i.e. throughput) and forget about this OLSR-like value (it would also be difficult to compute).

Actions #4

Updated by Sven Eckelmann over 7 years ago

  • Estimated time set to 40.00 h
Actions

Also available in: Atom PDF