Re: [LARTC] TCP window based shaping
TAKANO Ryousei wrote: Andy Furniss wrote: Don Cohen wrote: I recalled a discussion of manipulating outgoing tcp windows in order to control return tcp traffic. I finally found at least some of that discussion in a thread with the subject above (of this message). But I thought someone announced an implementation and I don't see it under this thread. If anyone else remembers or knows where I should look for it, please let me know. I'm also interested in other discussion of the idea and in implementations of related ideas such as delaying the acks etc. I don't think there is an implementation other than the commercial packeteer. Andy. We have been proposed TCP window based pacing (shaping) in the PFLDnet2005. You can get the paper and slides from the following URL: http://www.ens-lyon.fr/LIP/RESO/pfldnet2005/TechnicalProgram.php Looks intresting but I can't get the pdf (not found) and haven't got anything that opens .ppt s This software called PSPacer is avaiable. But, I am sorry, the current release version does not support TCP window based pacing whereas it supports static pacing. Am I right in thinking that PSPspacer has to be run on the server - in which case would the tcp window shaping also be for locally generated traffic only, or do both work on routers? Andy. ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
[LARTC] (no subject)
Hi. Need some solution to resolve Vlan on Fedora Core 2 release Operating system configured on HP Proliant DL 320 G3 server (Monitor Server). The system is part of telecommunication equipment to manage the data VLAN tagging of data or management traffic.I have enclosed a brief connectivity diagram. The problem is when FTP is initiated within VLAN the downloading is reduced to half, user is configured to 512 KBPS Broadband Connection. If the vlan is removed the FTP is fine both direct (Down Uploading). I feel the problem is somever in the VLAN configuration which we are not to pinpoint the problem. Please suggest!. Thanx in Advance! Regards Supratim Slide1.GIF Description: GIF image ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
Re: [LARTC] Simple traffic shaping
On Mon, 11 Jul 2005, gypsy wrote: My needs is limiting the outbound traffic of an smtp mail server. It is connected to a gateway via 100Mbit ethernet. I want limits its outbound traffic to max 3 Mbit. HTB: tc qdisc add dev $DEV root handle 1: htb default 20 Create the root and set the default for traffic to filter/class 20. Needed (filter/class 20)? tc class add dev $DEV parent 1: classid 1:1 htb rate 3000kbit burst 6k Create the class 1:1, set maximum rate to 3mbit. Can be useful increase the 6k burst? tc class add dev $DEV parent 1:1 classid 1:20 htb rate 3000kbit \ burst 6k quantum 1500 prio 1 ??? what do the quantum 1500 part? tc filter add dev $DEV parent 1: protocol ip prio 5 u32 \ match ip sport 25 0x flowid 1:20 Create a filter for smtp traffic? Why sport 25? I am interested in outgoing traffic. This box receive the outgoing mails from other internal servers and do the delivery. No other activity/traff. So can be useful that it receive traffic from local server at full speed, but delivery it at limited rate (the problem is mailing list users that sometime distribuite big mail (0.5-2 MB) to 1.000-3.000 subscribers causing peak that ... :-( Really I am thinking to use it to shape the total traffic from a server with no differentiation on services (so I can use it in mail or httpd server ...). Regards, B. ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
RE: [LARTC] HTB Rate and Prio (continued)
I had a go with what you posted there over lan and with 2 tcp streams it behaves as expected (see below for exact test). Can you reproduce the failiure shaping over a lan rather than your internet connection? If your upstream bandwidth is sold as 2meg then ceil 2000kbit is likely to be too high. You could also try specifying quantum = 1500 on all the leafs as you get it auto calculated from rates otherwise (you can see them with tc -s -d class ls ...). It didn't affect my test though. If you are looking at htbs rate counters remember that they use a really long average (about 100 sec) so they can mislead. I tested below with two netperf tcp send tests to ips I added to another PC on my lan. # /usr/local/netperf/netperf -H 192.168.0.60 -f k -l 60 /usr/local/netperf/netperf -f k -H 192.168.0.102 -l 60 which gave - Recv SendSend Socket Socket Message Elapsed Size SizeSize Time Throughput bytes bytes bytessecs.10^3bits/sec 43689 16384 1638460.091884.66 Recv SendSend Socket Socket Message Elapsed Size SizeSize Time Throughput bytes bytes bytessecs.10^3bits/sec 43689 16384 1638460.22 51.58 Did the exact same test and it's working (10kbits for the low prio was the only diff) !! That's with pfifo - TCP STREAM TEST to 10.0.1.228 Recv SendSend Socket Socket Message Elapsed Size SizeSize Time Throughput bytes bytes bytessecs.10^3bits/sec 87380 8192 819263.00 35.37 TCP STREAM TEST to 10.0.1.227 Recv SendSend Socket Socket Message Elapsed Size SizeSize Time Throughput bytes bytes bytessecs.10^3bits/sec 87380 8192 819260.001897.27 That's with sfq - TCP STREAM TEST to 10.0.1.227 Recv SendSend Socket Socket Message Elapsed Size SizeSize Time Throughput bytes bytes bytessecs.10^3bits/sec 87380 8192 819260.001918.02 TCP STREAM TEST to 10.0.1.228 Recv SendSend Socket Socket Message Elapsed Size SizeSize Time Throughput bytes bytes bytessecs.10^3bits/sec 87380 8192 819260.00 28.40 10.0.1.228 was the lowprio IP. So everything worked fine for the test...!!! After that I tested with my original script, I just changed the netfilter rules to classify packets with the same rules I used for the netperfs test and I had the same good results... So it must be something with my 2m line or with the traffic type I shape. I quite don't understand the concept of putting the rate of the line lower than it's true value, can you explain me this and do the excess bandwith is lost ? What is a good value for a 2m line (SDSL) ? I'll try tomorrow to have an host outside with netperf so I can test the line itself. Again thanks for your time and help Andy. ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
Re: [LARTC] TCP window based shaping
We have been proposed TCP window based pacing (shaping) in the PFLDnet2005. You can get the paper and slides from the following URL: http://www.ens-lyon.fr/LIP/RESO/pfldnet2005/TechnicalProgram.php Looks intresting but I can't get the pdf (not found) and haven't got anything that opens .ppt s Sorry! You can get the paper from the following URL: http://www.gridmpi.org/publications/pfldnet05-takano.pdf This software called PSPacer is avaiable. But, I am sorry, the current release version does not support TCP window based pacing whereas it supports static pacing. Am I right in thinking that PSPspacer has to be run on the server - in which case would the tcp window shaping also be for locally generated traffic only, or do both work on routers? I think tcp window based shaping is difficult on routers. PSPacer basically assumes the use of end-to-end traffic engineering. (PSPacer can also regulate bandwidth of through traffic at static target transmission rate on routers.) Thanks, TAKANO Ryousei ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
[LARTC] Teql and NetEm can't work together
Thanks in advance! Summary: when I load netem and teql together, teql doesn't work correctly. (If I load teql only, everything is fine) I loaded both netem and teql. Netem is associated with eth0, and teql is associated with both eth0 and eth1. But traffic only goes out of eth1.Attached are the commands that I used to configure teql and netem (on machine 1), and commands to configure teql only (on machien 2), their routing tables and "tc qdisc ls" results. What's wrong with my configuration? (I tried loading teql only without netem, and teql works asexpected: ping traffic goes to both interfaces. ) = Machine 1 with NetEm Commands to install NetEm and TEQL together: # modprobe sch_teql# tc qdisc add dev eth0 root handle 1: netem delay 10ms # tc qdisc add dev eth0 parent 1:1 handle 10: teql0 # tc qdisc add dev eth1 root teql0 # ip link set dev teql0 up # ip addr add dev eth0 10.0.0.1/30 # ip addr add dev eth1 10.0.0.5/30 # ip addr add dev teql0 10.0.0.9/30 # echo 0 /proc/sys/net/ipv4/conf/eth0/rp_filter# echo 0 /proc/sys/net/ipv4/conf/eth1/rp_filter [EMAIL PROTECTED] tc qdisc lsqdisc netem 1: dev eth0 limit 1000 delay 100.0ms qdisc teql0 10: dev eth0 parent 1:1 qdisc teql0 8001: dev eth1 qdisc pfifo_fast 0: dev teql0 bands 3 priomap 1 2 2 2 1 2 0 0 1 1 1 1 1 1 1 1 Destination Gateway Genmask Flags Metric Ref Use Iface10.0.0.8 0.0.0.0 255.255.255.252 U 0 0 0 teql010.0.0.0 0.0.0.0 255.255.255.252 U 0 0 0 eth010.0.0.4 0.0.0.0 255.255.255.252 U 0 0 0 eth1 == Machine 2 without NetEm == Commands to install TEQL: # modprobe sch_teql# tc qdisc add dev eth0 root teql0# tc qdisc add dev eth1 root teql0# ip link set dev teql0 up# ip addr add dev eth0 10.0.0.2/30# ip addr add dev eth1 10.0.0.6/30# ip addr add dev teql0 10.0.0.10/30# echo 0 /proc/sys/net/ipv4/conf/eth0/rp_filter# echo 0 /proc/sys/net/ipv4/conf/eth1/rp_filter [EMAIL PROTECTED] tc qdisc lsqdisc teql0 8001: dev eth0qdisc teql0 8002: dev eth1qdisc pfifo_fast 0: dev teql0 bands 3 priomap 1 2 2 2 1 2 0 0 1 1 1 1 1 11 1 Destination Gateway Genmask Flags Metric Ref Use Iface10.0.0.8 0.0.0.0 255.255.255.252 U 0 0 0 teql010.0.0.0 0.0.0.0 255.255.255.252 U 0 0 0 eth010.0.0.4 0.0.0.0 255.255.255.252 U 0 0 0 eth1 Best,-Ji ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
Re: [LARTC] QOS HELP PLEASE
ok i did the calculations and here it is : www.tdi.pozman.pl/fir3 now ping to world from server are superb but. ping to access points in the network are 1-2seconds... people keep on calling and tlling that tranfers on parts of network are 1-3kb/s. i hae to mention that on the old server all was perfect. please anyone :) -- *Dariusz 'tdi' Dwornikowski | Gentoo | admin at pozman.pl | *[JID]:[EMAIL PROTECTED]|[gg]:2266034|[IRC]:[EMAIL PROTECTED] | *[MAIL]:[EMAIL PROTECTED]|[WWW]:www.tdi.pozman.pl | *Serwery,administracja,webapps - www.ProAdmin.com.pl | *Fingerprint:43E21CC46DAFD2F754E91547D59B39F56AAA4B5F | pgp3PJXID7CuN.pgp Description: PGP signature ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
Re: [LARTC] QOS HELP PLEASE
On Tue, 2005-07-12 at 19:47 +0200, Dariusz Dwornikowski wrote: ok i did the calculations and here it is : www.tdi.pozman.pl/fir3 now ping to world from server are superb but. ping to access points in the network are 1-2seconds... people keep on calling and tlling that tranfers on parts of network are 1-3kb/s. i hae to mention that on the old server all was perfect. please anyone :) ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc Anytime I have ever had this problem which was 4 times 2 weeks ago was due to a loop in the network. I have no explanation for it and am pulling my hair out. Some time rebooting AP works sometimes reboot a switch works. Its not my network I am just trying to help someone fix it. Have not had the problem since I told the guy to pull all the cables out and re cable the network. I would not of said this only your iptables and tc looks like mine and the problem sounds so-similar. But this was the only post I could find from you. ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
[LARTC] htbinit Web manager
Hi. Anybody can tell me if exist na web interface to manage /etc/sysconfig/htb files ? That interface will help me so much creating and changing speed settings of my users. Thanks, André Ribeiro ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
Re: [LARTC] htbinit Web manager
Perhaps you wanna try this: http://shaper.netshadow.at/ Cheers, Andreas Andre Ribeiro wrote: Hi. Anybody can tell me if exist na web interface to manage /etc/sysconfig/htb files ? That interface will help me so much creating and changing speed settings of my users. Thanks, André Ribeiro ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
[LARTC] QOS problem -ng
ok i gave up. can sb point me docs on how design proper qos rules ? about differences between them and the usage (which for wifi, which for ISPs which for homenet) ill read and do it -- *Dariusz 'tdi' Dwornikowski | Gentoo | admin at pozman.pl | *[JID]:[EMAIL PROTECTED]|[gg]:2266034|[IRC]:[EMAIL PROTECTED] | *[MAIL]:[EMAIL PROTECTED]|[WWW]:www.tdi.pozman.pl | *Serwery,administracja,webapps - www.ProAdmin.com.pl | *Fingerprint:43E21CC46DAFD2F754E91547D59B39F56AAA4B5F | pgpElb5on6XAu.pgp Description: PGP signature ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
[LARTC] Two class C subnets...
Hi!, I'm getting nuts with this so I really need your help... I have a network in the 192.168.100.0/24 class C with the gw in 192.168.100.1, all the company works great with that scenario but now we need to add more class C networks, now this is what I have: 2: eth0: BROADCAST,MULTICAST,UP mtu 1500 qdisc pfifo_fast qlen 1000 link/ether 00:08:a1:53:6d:51 brd ff:ff:ff:ff:ff:ff inet 192.168.100.1/24 brd 192.168.100.255 scope global eth0 inet 192.168.2.1/24 brd 192.168.2.255 scope global eth0 0: from all lookup local 32763: from 192.168.100.0/24 to 192.168.2.0/24 lookup Monitors2 32764: from 192.168.2.0/24 to 192.168.100.0/24 lookup Monitors 32765: from xxx.xxx.xxx.xxx lookup ISP 32766: from all lookup main 32767: from all lookup 253 Monitors Table: default via 192.168.100.1 dev eth0 Monitors2 Table: default via 192.168.2.1 dev eth0 Main Table: 192.168.100.0/24 dev eth0 scope link 192.168.2.0/24 dev eth0 proto kernel scope link xxx.xxx.xxx.xxx/24 dev eth2 scope link 127.0.0.0/8 dev lo scope link default via xxx.xxx.xxx.xxx dev eth2 The only thing that I want is that if a packet comming from the 192.168.2.0/24 is going to some host in the 192.168.100.0/24 net, the kerenl route those packets to 192.168.100.1 or in the other way, now, I don't know what I'm doing wrong but that doesn't work for me, the packets got lost somewhere, I'm a little lost with this so please can anyone give me a hint? Thanks -- Nicolas A. Salvo Capital Federal Buenos Aires - Argentina ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
[LARTC] tbf initial burst
I am using tbf to do bandwidth limitation. i found that when i start passing traffic there is a burst and then the rate goes down to what is configured. is this a known issue or do i need to change some parameters? thanks. Do you Yahoo!? Read only the mail you want - Yahoo! Mail SpamGuard.___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
Re: [LARTC] tbf initial burst
Greetings, : I am using tbf to do bandwidth limitation. i found that when i : start passing traffic there is a burst and then the rate goes : down to what is configured. is this a known issue or do i need to : change some parameters? The behaviour you have described is exactly the theoretical goal of a token bucket filter, and also the practical behaviour of the TBF queueing discipline. In other words, congratulations, you are using a TBF! You probably wish to tweak parameters. -Martin -- Martin A. Brown --- Wonderfrog Enterprises --- [EMAIL PROTECTED] ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
[LARTC] HTB: ¿how do burst/cburst work exactl y?
I´ve read all the definitions of burst and cburst and I´ve tried playing with the parameters and graphing the output of the filter to see its effects, but STILL I can´t figure out how the parameters work exactly. ¿Could anyone give a better explanation than the manpage? ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
Re: [LARTC] HTB: ¿how do burst/cburst work exactl y?
Hello, : I´ve read all the definitions of burst and cburst and I´ve tried : playing with the parameters and graphing the output of the filter : to see its effects, but STILL I can´t figure out how the : parameters work exactly. : : Could anyone give a better explanation than the manpage? Have you tried Stef's site? He has a good page [0] that talks about the various tests he did while experimenting with HTB burst and cburst parameters? Some time ago, I took a stab at creating a visual representation [1] of a hypothetical HTB configuration [2]. In order to understand when cburst is used, look for the diamond-shaped boxes in the diagram which talk about tokens and ctokens. Every HTB class has two buckets. rate bucket is of burst size, traffic uses tokens ceil bucket is of cburst size, traffic uses ctokens My diagram may give you the framework to understand exactly how they are used if it's still unclear to you, but Stef's site will give you much better detail on the results of using burst and cburst. Of the scenarios he describes, I like the results of Test 7 the best. The only guideline that struck me after reading his results was to prefer burst and cburst usage on parent classes. Good luck, -Martin [0] http://www.docum.org/docum.org/tests/htb/burst/ [1] http://linux-ip.net/traffic-control/htb-class.png [2] http://linux-ip.net/traffic-control/diagram.html -- Martin A. Brown --- Wonderfrog Enterprises --- [EMAIL PROTECTED]___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc
Re: [LARTC] QOS problem -ng
Dariusz Dwornikowski wrote: ok i gave up. can sb point me docs on how design proper qos rules ? about differences between them and the usage (which for wifi, which for ISPs which for homenet) ill read and do it Start with Jim diGriz in shaping / QoS / Traffic Control at http://yesican.chsoft.biz/lartc Then read Dan Singletary and Emmanuel Roger. That ought to get you going. -- gypsy ___ LARTC mailing list LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc