Adis Nezirovic wrote:
after incoming packets jump to ( -i eth0 -j IMQ1) i made another jump when packets leaving out eth1 (-o eth1 -j IMQ1) he he he , results, kernel crash and reboot several times when big packets arrived. So, enough for the try n error, wont try again.I think kernel panics are triggered by IMQ. I was able to solve that problem with following rules: iptables -t mangle -A POSTROUTING -o eth0 -s ! 192.168.100.100 -j IMQ --todev imq0 iptables -t mangle -A PREROUTING -i eth0 -j IMQ --todev imq0 (eth0 is my internal NIC, with 192.168.100.100 ip address). i.e. no download shaping from "local machine". By using the same IMQ and NIC device for both, POSTROUTING and PREROUTING, I am able to simulate "half-duplex" link. (bandwidth for upload/download is shared). tcng works great in the above combination. Kernel 2.6.11.11 (with ck patches)
I am lucky I don't really have to shape for single duplex.There are alternatives eg if you only need to do forwarded traffic mark -i eth0 -o eth1 in forward and visa versa and then jump to imq for the marked packets just from postrouting.
If you need to include local traffic and don't need to hook imq after de-nat in prerouting you could use a modified dummy device instead.
Replace drivers/net/dummy.c with the one attached rebuild and you can shape with it.
I can't find the script I tested but could do another example sometime if anyone needed.
Andy.
Attachment:
dummy.c.gz
Description: GNU Zip compressed data
_______________________________________________ LARTC mailing list LARTC@xxxxxxxxxxxxxxx http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
