Difference between revisions of "Mesh/Interviews with other meshers"

no edit summary
(Added Altermundi interview excerpts)
Line 1: Line 1:
= Interesting questions to ask =
= Interesting questions to ask =


Line 29: Line 28:


*How do you handle bandwidth shaping / quality of service on internet gateway nodes?
*How do you handle bandwidth shaping / quality of service on internet gateway nodes?
= Notes from talking to personal telco =
Notes from talking with scott garman at sudo room on May 15th 2014.
They don't limit internet sharing and have had very few issues.
They don't VPN connections (except for IPv6 connections through hurricane electric's free tunnel thing).
They use OpenVPN and perl on the nodes. This means they have some trouble running on only 32 mb of ram.
They mostly use OLSR but some nodes use batman-adv in small clusters.
They don't have very many linked nodes. Mostly they're just enabling internet sharing. One of their problems is flat terrain and many tall trees. They've used 900 mhz in some locations to better pass through foliage.
They haven't had any issues with abuse. They use MAC blocks on the node that is experiencing problems to block the problematic client from that specific node only. They've only had two instances were they client repeatedly changed their MAC to re-gain access and they just kept banning the new MAC addresses until the users gave up.
They haven't had any issues with Comcast with regards to the sharing of internet access.
As I understand it, they seem to configure the nodes for the specific scenario required. E.g. some nodes are just plain access points and some point to point nodes don't run the meshing protocols and some small clusters run a different meshing protocol.
They use cacti/munin + snmpd on the nodes for monitoring. They also have the nodes transmit their log files to a central server.


= Seattle Meshnet Interviews =
= Seattle Meshnet Interviews =