[Babel-users] Reachability issue when multicast is less reliable than unicast

[StarBrilliant]

Hi there,

I run a babeld cluster over ZeroTier mesh VPN with 10 nodes spread
globally. Recently I met some reachability issues when I add new nodes to
the network.

The kernel routing table periodically shows some nodes as "unreachable". I
made sure that missing nodes are in fact reachable via a link-local ping.
Also when I do a "ping6 ff02::1:6%zt0" over the network, sometimes I only
receive 7 or 8 replies. I guess ZeroTier has performance issues with
multicast (it says it can only run multicast over 32 nodes by default).

The following shows my current configuration, with hop count, packet drop,
and latency taken consideration into metrics calculation:

> interface zt0 type tunnel link-quality true rxcost 16 hello-interval 10
rtt-min 16 rtt-max 1024 max-rtt-penalty 1008
> redistribute local if zt0 allow
> redistribute local deny

I know Babel can now run over unicast addresses. How can I improve my
babeld configuration to utilize the new feature?

(P.S. To my knowledge, multicast over Wi-Fi has lower transmission rate but
no retry, thus also has different reliability compared to unicast. How much
would babel's performance be affected over such networks?)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://alioth-lists.debian.net/pipermail/babel-users/attachments/20180420/78fdb8a2/attachment.html>