similar to: Best way to manage a 128Kb/s pipe - help

Hi, I am trying to do both ingress and egress bandwidth management on a bridge. eth0 is the WAN interface. IPTraf shows that the following script is running successfully and it limits bandwidth both ways to 256 KB However when I want to see the statistics using tc - it only shows me stats for htb class and not for ingress class. The tc output is as follows: tc -s -d class show dev eth0 class

Hi All, I read the advanced networking howto and find a useful way to shape my adsl link. is says: "tc qdisc add dev ppp0 root tbf rate 220kbit latency 50ms burst 1500 [...] Change 220kbit to your uplinks *actual* speed, minus a few percent. If you have a really fast modem, raise ''burst'' a bit. " My adsl''s upload brandwidth is about 120 kbits/sec,

Hello all. In the shower the other day (where most of this sort of musing gets done, eh?) I was thinking about graceful degradation of audio signals. Let me apologise in advance if these are elementary concepts or if I demonstrate a complete lack of insight -- I don't rate even a dabbler status in the area of audio codecs. Anyway: If we have a 128kbs signal coming down a *udp* channel with

Hello all, this is my first post here. Sorry for my english. Gentoo LAN router, 2.4.26-hardened-r2 There are 2 WAN links, one LAN link. I am doing some iptables/routing/tc magic in my scripts. What''s interesting is marking packets traveling from all IP''s in LAN. Interesting commands are: ------------- for ip in `seq 50`; do $IPTABLES -t mangle -A FORWARD -o eth2 -d

ok .. a use iptraf, but iptraf show my evry connection from computerX: 84.135.15.81:4899 > 163 7940 --A- eth0 ? 82.77.210.7:3475 > 212 323323 -PA- eth0 ? 62.231.96.133:80

Hi, I''ve been using htb fine to separate outgoing SMTP traffic from HTTP traffic so sending big emails doesn''t affect browsing. As long as ceil == burst and the sum of all rates don''t exceed the root class'' rate all is fine but when I incorporate ceil I would expect (from lartc howto) that when other classes are idle the remaining bandwith would be assigned

Hi everyone, My HTB script is showing me those "r2q change" warnings. I read that I should test with r2q values to gain accuracy and stop the warnings but I made some tests with r2q and with burst and I still get the warnings. Any ideas on this issue will be very helpful. The script seems to be working as the bandwidth for the IPs is being shaped. The NIC works at 100mbit and the

I have a CentOS 4 server at home where we store all our data, I wanted to get some idea of the amount of traffic going into and out of the server by protocol so I installed IPtraf.. When I run it from an ssh session it looks fine but when I run it directly on the server console its doesn't look right at all.. It doesn't seem to layout correctly.. Is there something I need to do

This is one of my test classes: class htb 1:10 parent 1:1 prio 0 quantum 2048 rate 160Kbit ceil 400Kbit burst 1803b/8 mpu 0b cburst 2111b/8 mpu 0b level 0 Q1: where does "level 0" stand for? Q2: where does this b/8 stand for? Q3: this is on a i386 platform, so timer resolution should be 10mS. According to the doc the minimal burst should be 10mS*160Kbit=1600. Why is it 1803? Q4: I

I am using HTB+IMQ on an Ethernet bridge and tc for statistics reporting. I am finding that tc reported rate is about 5-10% less than that reported by IPTraf for the same traffic. Has anyone observed this before? Which is more accurate? Rajesh

Dear All, Sorry to trouble again..... After go through www.lartc.org I have implemented the HTB instead of CBQ for the same scenario. Now following files are under /etc/sysconfig/htb directory. eth0 DEFAULT=30 R2Q=10 eth0-2.root RATE=256kbps BURST=25k eth0-2:10.comp1 RATE=120kbps BURST=12k PRIO=0 LEAF=sfq RULE=192.168.200.0/24 eth0-2:20.comp2

Hi I have setup iproute2 and need a bit of help reading the stats from it ===== output qdisc htb 1: r2q 10 default 20 direct_packets_stat 0 ver 3.17 Sent 547326809 bytes 1342627 pkts (dropped 9303, overlimits 2817572 requeues 0) backlog 46p qdisc sfq 10: limit 128p quantum 1514b flows 128/1024 perturb 10sec Sent 41874343 bytes 730889 pkts (dropped 0, overlimits 0 requeues 0) qdisc sfq

All: Does anyone know what the burst and cburst parameter do? My understanding so far: * I see a lot of different definitions on the web. It seems like burst is the number of bytes sent before serving other queues/classes. So if burst was 1000 bytes and class rate was 100kibit per second. It would send 1000 bytes each time the scheduler service that queue to a rate of 100 kbit per

//Just to not forget eth0 is WAN and eth1 is LAN . The box is NATing my lan. I configured tcng to shape both download and upload and i got connection loss again. Maybe it is my script fault or it is something bad with packets marking . this is how i mark outgoing packets iptables -t mangle -A MYSHAPER-OUT -p tcp -s 192.168.1.23 -j MARK --set-mark $ip and this part of my script which is intented

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, I was just measuring the bitrates of a couple of codecs via iax. I'm getting much higher numbers than expected, so maybe I'm doing something wrong? Measured with iptraf, values displayed are: codec: measured bitrate (bitrate according codec definition) gsm: 52 kbps (13 kpbs) alaw: 154 kbps (?) speex: 57 kpbs (24 kpbs) Seems a little

I have htb on 486 sx with 16mb ram. Slackware 9.1. Connection : dsl 1Mbit. 486 works as router and trafic shaper for network made of 12 pc''s. it does the job quite well, but when i play Counter-Strike or connect to my 486 via ssh (on lan), i get huge lag every 11-20 sec. when i connect to 486 via ssh and run iptraf program i see all the trafic, and after a while when lag comes the

I''m trying to fill a token bucket with enough tokens to burst several gigs of data. However, it doesn''t seem to get any higher than ~3.9GB: >tc qdisc add dev eth0 root tbf rate 1440kbit latency 50ms \ burst 16000000000 >tc qdisc show dev eth0 qdisc tbf 800b: rate 1440Kbit burst 3908420240b lat 2197.8s A smaller attempt of ~1.6 gigs works just fine: >tc qdisc

I currently have Icecast 2.3.1 running with Darkice on a Fedora Core 4 box (with the 2.6.14 kernel). We have three streams running (all mp3) @ 128kbs, 64kbs, and 24kbs. I am consistently seeing the following error: kernel: Unable to handle kernel paging request at virtual address ffffffff It does appear to be icecast causing the error. Oddly, when this first started showing up, it would

Hi! I am studying HTB. I used the topology showed below: 10.1.0.1 ------ 10.1.14.25 (eth1) / 192.168.10.10 (eth0) ----- 192.168.10.11 (host 1) (router) (host 2) All machines use Debian Etch. Has a Iptables masquerading rule to eth1 in the router machine . The NICs are 100 Mb/s. The host 1 has Apache 2 and a file with 670 MB (CD ISO image) to download. When I used