NetBalancer drops every packet

[Istaria]

Hey,

I’ve been working on all sorts of problems for almost a week now and am getting pretty frustrated and desperate. Hence this post.

Here’s the setup from zs’s view:

LAN network 192.168.0.x with 6 computers on eth0
ADSL network 83.245.x.x on eth1
LAN network 192.168.1.x with 2 computers on eth2

192.168.1.1 is zeroshell eth2
192.168.1.2 is a computer with a 3g modem

If I hook up any computer to the 192.168.1.x network and set the gateway to be the 3g modem computer, I can connect to internet just fine. zs failover ip address test gives me this response:

+++ Testing across the gateway DEFAULT GATEWAY (Gateway: 83.245.x.x) +++
[188.40.52.167] : Success (Round Trip Time: 44.4 ms)
[193.110.226.14] : Success (Round Trip Time: 11.3 ms)
[193.65.255.152] : Success (Round Trip Time: 11.8 ms)
+++ Testing across the gateway WARETUS GATEWAY (Gateway: 192.168.1.2) +++
Failover monitor has started
[188.40.52.167] : Success (Round Trip Time: 820 ms)
[193.110.226.14] : Success (Round Trip Time: 445 ms)
[193.65.255.152] : Success (Round Trip Time: 425 ms)

…which (together with earlier testing) indicates that both gateways are set up at least somewhat correctly. However, when I try to set up netbalancer to use either gateway, I won’t get past the zeroshell computer.

Here’s a log entry of a force-netbalanced packet that won’t get anywhere:

NetBalancer/001IN=ETH00 OUT= MAC=00:0c:f1:7d:f7:4b:00:11:09:cc:96:07:08:00 SRC=192.168.0.1 DST=82.103.135.97 LEN=56 TOS=0x00 PREC=0x00 TTL=8 ID=26380 PROTO=ICMP TYPE=8 CODE=0 ID=1024 SEQ=54813

Activity is registered on eth2 but not on 192.168.1.2’s network interface.

Routing table is as follows (edited for readibility):
192.168.0.0 – 255.255.255.0 – Net – 0 – none – ETH00 – U – Up – Auto
192.168.1.0 – 255.255.255.0 – Net – 0 – none – ETH02 – U – Up – Auto
83.245.x.0 – 255.255.248.0 – Net – 0 – none – ETH01 – U – Up – Auto
+ default gateways

Gateway configuration is:
Default gateway – weight 99 – ip addr 83.245.x.x – interface none – timeout 1
Second gateway – weight 1 – ip addr 192.168.1.2 – interface none – timeout 8

If I missed something relevant, I’ll be happy to post. Thank you in advance for any help. I’m a linux newb and it has been quite hard to get this far. I’m thinking I’m already pretty close but sadly no cigar.

[atheling]

First, do you have my patch installed? You can get the version for b14 at: http://dl.dropbox.com/u/19663978/ZS_nb_quo_b14_b.tar.gz

Second, it would be useful to see the output from the ZS console of the following shell commands:

ip rule list
ip route list table 101

(repeat the above for each of lines that end in “lookup nnn” using the “nnn” value listed.)

ip route list main
iptables -L -v -n -t mangle

This should show how the GUI actually setup the Linux routing and iptables used in netbalancing and QoS. Assuming you don’t actually have a iptables forward rule prohibiting the packets this should help figure out what went wrong.

[Istaria]

Thank you for the response! I do/did indeed have b14.

I just got a network expert friend of mine to configure everything remotely. “Sadly” it works now, and thus I can’t help anyone else with what the solution was as I have no idea what was wrong originally.

Now if only the 3g modem didn’t drop the connection occasionally without realizing that it has dropped… looks like I got to make some sort of a script that finds out if the connection is still working and if it isn’t, identifies the pppd’s pid and kills it which causes wvdial on the 3g gateway ubuntu machine to finally figure out something is wrong and to redial.

For some reason, only torrent data seems to correctly route through gateway 2 using netbalancing rules. Other LAYER7 configurations don’t seem to do anything just yet so I guess there’s still a whole lot of things for me to do.

A bit of basic linux knowledge (even if it’s all gotten from google and probably instantly forgotten) won’t do any harm so a nice and fun project to work on 🙂

[Author]

Posts