tc class add dev eth1 parent 1: classid 1:1 htb rate 10kbps ceil 10kbps burst 2k why the client run into 1:1 can keep 430 KB/s ? PC+ patch HTB + linux2.4.16
On Tue, Dec 11, 2001 at 10:46:08AM +0800, yangrunhua wrote:> tc class add dev eth1 parent 1: classid 1:1 htb rate 10kbps ceil 10kbps > burst 2k > why the client run into 1:1 can keep 430 KB/s ?Show your entire configuration, please. Regards, bert hubert -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet
tc qdisc add dev eth0 root handle 1: htb default 2 tc class add dev eth0 parent 1: classid 1:1 htb rate 5kbps ceil 5kbps burst 2k tc class add dev eth0 parent 1: classid 1:2 htb rate 10mbps ceil 10mbps burst 2k tc filter add dev eth0 protocol ip fw and use iptables to mark every packet 0x10001,but up to 200KB/s. # tc -s -d qdisc qdisc htb 1: dev eth0 r2q 10 default 2 dcache 0 deq_util 1/1000000 deq_rate 10 trials_per_deq 0 dcache_hits 0 Sent 1213554 bytes 13996 pkts (dropped 0, overlimits 3924) # tc -s -d class show dev eth0 class htb 1:1 root prio 0 rate 40Kbit ceil 40Kbit burst 2Kb/8 mpu 0b cburst 2Kb Sent 209473 bytes 3364 pkts (dropped 0, overlimits 3367) rate 73bps 1pps lended: 3364 borrowed: 0 giants: 0 injects: 0 tokens: 316161 ctokens: 316161 class htb 1:2 root prio 0 rate 80Mbit ceil 80Mbit burst 2Kb/8 mpu 0b cburst 2Kb Sent 1011846 bytes 10712 pkts (dropped 0, overlimits 151) rate 258bps 1pps lended: 10712 borrowed: 0 giants: 0 injects: 0 tokens: 155 ctokens: 155 why? PII400 + 64M + 2.4.16 -----Original Message----- From: lartc-admin@mailman.ds9a.nl [mailto:lartc-admin@mailman.ds9a.nl] On Behalf Of bert hubert Sent: 2001年12月11日 17:21 To: lartc@mailman.ds9a.nl Subject: Re: [LARTC] HTB problem On Tue, Dec 11, 2001 at 10:46:08AM +0800, yangrunhua wrote:> tc class add dev eth1 parent 1: classid 1:1 htb rate 10kbps ceil10kbps> burst 2k > why the client run into 1:1 can keep 430 KB/s ?Show your entire configuration, please. Regards, bert hubert -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet _______________________________________________ LARTC mailing list / LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://ds9a.nl/lartc/
Sorry,forget sth. Ifconfig sees: RX bytes:22970234 (21.9 Mb) TX bytes:1741892 (1.6 Mb) But: # tc -s -d qdisc qdisc htb 1: dev eth0 r2q 10 default 2 dcache 0 deq_util 1/1000000 deq_rate 11 trials_per_deq 0 dcache_hits 0 Sent 1301198 bytes 14448 pkts (dropped 0, overlimits 3944) Why? -----Original Message----- From: lartc-admin@mailman.ds9a.nl [mailto:lartc-admin@mailman.ds9a.nl] On Behalf Of bert hubert Sent: 2001年12月11日 17:21 To: lartc@mailman.ds9a.nl Subject: Re: [LARTC] HTB problem On Tue, Dec 11, 2001 at 10:46:08AM +0800, yangrunhua wrote:> tc class add dev eth1 parent 1: classid 1:1 htb rate 10kbps ceil10kbps> burst 2k > why the client run into 1:1 can keep 430 KB/s ?Show your entire configuration, please. Regards, bert hubert -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet _______________________________________________ LARTC mailing list / LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://ds9a.nl/lartc/
On Tue, Dec 11, 2001 at 05:34:22PM +0800, yangrunhua wrote:> tc qdisc add dev eth0 root handle 1: htb default 2 > tc class add dev eth0 parent 1: classid 1:1 htb rate 5kbps ceil 5kbps > burst 2k tc class add dev eth0 parent 1: classid 1:2 htb rate 10mbps > ceil 10mbps burst 2k > tc filter add dev eth0 protocol ip fwI still have a very hard time figuring out your configuration. Please list the following: * your iptables commands * your tc qdisc add commands * your tc class add commands * what kind of traffic you are generating Thanks. -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet
Now, I know the where the problem is. If I already use SNAT, how could I limit the inbound bandwidth of Every intra-network ip protected by SNAT( with the help of iptables nfmark)? But mangle at PREROUTING goes before NAT translate. Is there any method? Thanks. -----Original Message----- From: bert hubert [mailto:ahu@ds9a.nl] Sent: 2001年12月11日 19:10 To: yangrunhua Cc: lartc@mailman.ds9a.nl Subject: Re: [LARTC] HTB problem On Tue, Dec 11, 2001 at 05:34:22PM +0800, yangrunhua wrote:> tc qdisc add dev eth0 root handle 1: htb default 2 > tc class add dev eth0 parent 1: classid 1:1 htb rate 5kbps ceil 5kbps> burst 2k tc class add dev eth0 parent 1: classid 1:2 htb rate 10mbps > ceil 10mbps burst 2k > tc filter add dev eth0 protocol ip fwI still have a very hard time figuring out your configuration. Please list the following: * your iptables commands * your tc qdisc add commands * your tc class add commands * what kind of traffic you are generating Thanks. -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet
I could manage it now, modified the netfilter src, add another table not mangle, Which runs after NAT_SRC. :-) Another problem: We know HTB is a tree structure, does it permit to jump to the leaf directly?(I use cls_fw which attached to the root, and with no options, so cls_fw would use nfmark of skb as classid and jump directly into leaf, no traverse) but when I use these: tc qdisc add dev eth0 root handle 1: htb default 12 tc class add dev eth0 parent 1: classid 1:1 htb rate 100kbps ceil 100kbps burst 2k tc class add dev eth0 parent 1:1 classid 1:10 htb rate 30kbps ceil 100kbps burst 2k tc class add dev eth0 parent 1:1 classid 1:11 htb rate 10kbps ceil 100kbps burst 2k tc class add dev eth0 parent 1:1 classid 1:12 htb rate 60kbps ceil 100kbps burst 2k tc filter add dev eth0 protocol ip fw that, packets marked with 0x1000a don''t jump into 1:10, why? But when I use these: tc qdisc add dev eth0 root handle 1: htb default 2 tc class add dev eth0 parent 1: classid 1:1 htb rate 5kbps ceil 5kbps burst 2k tc class add dev eth0 parent 1: classid 1:2 htb rate 10kbps ceil 10kbps burst 2k tc filter add dev eth0 protocol ip fw in this situation, packets marked with 0x10001 succeeded in jumping into 1:1. Why? -----Original Message----- From: lartc-admin@mailman.ds9a.nl [mailto:lartc-admin@mailman.ds9a.nl] On Behalf Of yangrunhua Sent: 2001年12月12日 13:47 To: lartc@mailman.ds9a.nl Subject: RE: [LARTC] HTB problem Now, I know the where the problem is. If I already use SNAT, how could I limit the inbound bandwidth of Every intra-network ip protected by SNAT( with the help of iptables nfmark)? But mangle at PREROUTING goes before NAT translate. Is there any method? Thanks. -----Original Message----- From: bert hubert [mailto:ahu@ds9a.nl] Sent: 2001年12月11日 19:10 To: yangrunhua Cc: lartc@mailman.ds9a.nl Subject: Re: [LARTC] HTB problem On Tue, Dec 11, 2001 at 05:34:22PM +0800, yangrunhua wrote:> tc qdisc add dev eth0 root handle 1: htb default 2 > tc class add dev eth0 parent 1: classid 1:1 htb rate 5kbps ceil 5kbps> burst 2k tc class add dev eth0 parent 1: classid 1:2 htb rate 10mbps > ceil 10mbps burst 2k > tc filter add dev eth0 protocol ip fwI still have a very hard time figuring out your configuration. Please list the following: * your iptables commands * your tc qdisc add commands * your tc class add commands * what kind of traffic you are generating Thanks. -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet _______________________________________________ LARTC mailing list / LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://ds9a.nl/lartc/
try replace 0x0001000a by 0x00010010. IIRC in 1:10 the "10" is in hex. devik On Wed, 12 Dec 2001, yangrunhua wrote:> I could manage it now, modified the netfilter src, add another table not > mangle, > Which runs after NAT_SRC. :-) > Another problem: We know HTB is a tree structure, does it permit to jump > to the > leaf directly?(I use cls_fw which attached to the root, and with no > options, so cls_fw would use nfmark of skb as classid and jump directly > into leaf, no traverse) > > but when I use these: > tc qdisc add dev eth0 root handle 1: htb default 12 > tc class add dev eth0 parent 1: classid 1:1 htb rate 100kbps ceil > 100kbps burst 2k > tc class add dev eth0 parent 1:1 classid 1:10 htb rate 30kbps ceil > 100kbps burst 2k > tc class add dev eth0 parent 1:1 classid 1:11 htb rate 10kbps ceil > 100kbps burst 2k > tc class add dev eth0 parent 1:1 classid 1:12 htb rate 60kbps ceil > 100kbps burst 2k > tc filter add dev eth0 protocol ip fw > > that, packets marked with 0x1000a don''t jump into 1:10, why? > But when I use these: > tc qdisc add dev eth0 root handle 1: htb default 2 > tc class add dev eth0 parent 1: classid 1:1 htb rate 5kbps ceil 5kbps > burst 2k > tc class add dev eth0 parent 1: classid 1:2 htb rate 10kbps ceil 10kbps > burst 2k > tc filter add dev eth0 protocol ip fw > > in this situation, packets marked with 0x10001 succeeded in jumping into > 1:1. > > Why? > > -----Original Message----- > From: lartc-admin@mailman.ds9a.nl [mailto:lartc-admin@mailman.ds9a.nl] > On Behalf Of yangrunhua > Sent: 2001Äę12ÔÂ12ČŐ 13:47 > To: lartc@mailman.ds9a.nl > Subject: RE: [LARTC] HTB problem > > Now, I know the where the problem is. > If I already use SNAT, how could I limit the inbound bandwidth of > Every intra-network ip protected by SNAT( with the help of iptables > nfmark)? > But mangle at PREROUTING goes before NAT translate. > Is there any method? > Thanks. > > > -----Original Message----- > From: bert hubert [mailto:ahu@ds9a.nl] > Sent: 2001Äę12ÔÂ11ČŐ 19:10 > To: yangrunhua > Cc: lartc@mailman.ds9a.nl > Subject: Re: [LARTC] HTB problem > > On Tue, Dec 11, 2001 at 05:34:22PM +0800, yangrunhua wrote: > > tc qdisc add dev eth0 root handle 1: htb default 2 > > tc class add dev eth0 parent 1: classid 1:1 htb rate 5kbps ceil 5kbps > > > burst 2k tc class add dev eth0 parent 1: classid 1:2 htb rate 10mbps > > ceil 10mbps burst 2k > > tc filter add dev eth0 protocol ip fw > > I still have a very hard time figuring out your configuration. Please > list > the following: > > * your iptables commands > * your tc qdisc add commands > * your tc class add commands > * what kind of traffic you are generating > > Thanks. > -- > http://www.PowerDNS.com Versatile DNS Software & Services > Trilab The Technology People > Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - > ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet > > > _______________________________________________ > LARTC mailing list / LARTC@mailman.ds9a.nl > http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: > http://ds9a.nl/lartc/ > > > _______________________________________________ > LARTC mailing list / LARTC@mailman.ds9a.nl > http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://ds9a.nl/lartc/ > >
On Wed, Dec 12, 2001 at 12:40:57PM +0100, Martin Devera wrote:> try replace 0x0001000a by 0x00010010. IIRC in 1:10 > the "10" is in hex.True. Even the 1 is hex! Regards, bert -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet
Thanks to everyone, my problem has been solved. Here is my description: for BIG ISP, if you have broadband for every client and use NAT_SRC to let them out, and you want to divide bandwidth among all clients. My solution is iproute2 + netfilter. I''ve modified netfilter framework''s src to hash every client (not like originally,line by line) and also do some accounting based on flow, and add a table(which hacks after NAT_SRC, not like mangle. Mangle hacks before NAT_SRC), then mark the packets with it''s classid related with it''s intra-network ipaddr, so the dev attached with filter fw with no option will classify directly into the class with that classid, finally control it''s bandwidth(I use HTB). Every step is O(1) in complexity. Excellent! Thanks for all. Thanks for all the people here and from netfilter mail-list. Iproute2+ netfilter seems pretty good! I''ve got good performance! I''m a new here, so very nice to hear from you, you are so kind. -----Original Message----- From: lartc-admin@mailman.ds9a.nl [mailto:lartc-admin@mailman.ds9a.nl] On Behalf Of bert hubert Sent: 2001年12月12日 21:51 To: Martin Devera Cc: yangrunhua; lartc@mailman.ds9a.nl Subject: Re: [LARTC] HTB problem On Wed, Dec 12, 2001 at 12:40:57PM +0100, Martin Devera wrote:> try replace 0x0001000a by 0x00010010. IIRC in 1:10 > the "10" is in hex.True. Even the 1 is hex! Regards, bert -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet _______________________________________________ LARTC mailing list / LARTC@mailman.ds9a.nl http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://ds9a.nl/lartc/
I have found CBQ to be very inaccurate (as others on the list)
I have created a structure as follow
Root (10mbit)
|
CBQ class (bandwidth=1mbit, rate=10mbit)
|
CBQ qdisc (bandwidth=1mbit)
Now when i send excess data to the tune of 1.3mbit, cbq lets it thru.
If i increase it beyond that, CBQ suddenly wakes up and throttles it down
to about 0.8Mbit!
Are there any explanations? If i am sending fixed size packets, should i
set the ''avpkt'' to that size?
I havent even able to use HTB correctly since a small burst size kills my
thruput.
Any ideas about CBQ/HTB or anything which will throttle my bandwidth to
1Mbit accurate *and* allow me to create qdiscs within. (Hence TBF is ruled
out). My HTB script is given below i am no clue why it doesnt work.
-----------------------------------------------------------------------
BANDWIDTH="bandwidth 10Mbit"
LIMITBW="1Mbit"
$TC qdisc add $DEVICE root handle 1: htb
# This HTB class rate limits everyting to 1Mbit
$TC class add $DEVICE parent 1: classid 1:1 htb rate $LIMITBW ceil
$LIMITBW burst 2k
----------------------------------------------------------------------
--
The statement below is true.
The statement above is false.
^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^
Amit Kucheria
EECS Grad. Research Assistant
University of Kansas @ Lawrence
(R)+1-(785)-830 8521 ||| (O)+1-(785)-864 7774
____________________________________________________
On Wed, Dec 12, 2001 at 07:17:43PM -0600, Amit Kucheria wrote:> I have found CBQ to be very inaccurate (as others on the list) > > I have created a structure as follow > > Root (10mbit) > | > CBQ class (bandwidth=1mbit, rate=10mbit) > | > CBQ qdisc (bandwidth=1mbit)Um, this doesn''t make sense. I keep repeating this to posters here, SHOW YOUR CONFIGURATION! Don''t just draw pictures - your CBQ commandlines may not do what you think they do. In this case, it appears you''ve attached a qdisc within a qdisc or something, I''m not sure. That could cause huge inaccuracies. So show your configuration. Regards, bert -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet
On Thu, 13 Dec 2001, bert hubert wrote:> On Wed, Dec 12, 2001 at 07:17:43PM -0600, Amit Kucheria wrote: > > I have found CBQ to be very inaccurate (as others on the list) > > > > I have created a structure as follow > > > > Root (10mbit) > > | > > CBQ class (bandwidth=1mbit, rate=10mbit) > > | > > CBQ qdisc (bandwidth=1mbit) > > Um, this doesn''t make sense. I keep repeating this to posters here, SHOW > YOUR CONFIGURATION! Don''t just draw pictures - your CBQ commandlines may not > do what you think they do.My apologies. Here''s the script --------------------------------------------------------------------------- BANDWIDTH=bandwidth 10Mbit LIMITBW=1Mbit # Root CBQ qdisc 1: $TC qdisc add $DEVICE root handle 1: cbq $BANDWIDTH $AVPKT # Root CBQ class 1:1 # This class rate limits everyting to 1Mbit $TC class add $DEVICE parent 1:0 classid 1:1 cbq $BANDWIDTH rate $LIMITBW \ maxburst 100 $AVPKT allot 1514 weight 100Kbit prio 1 bounded isolated # Filter packet from the 2 sources $TC filter add $DEVICE parent 1: protocol ip prio 1 u32 match ip src $SRC1 flowid 1:1 $TC filter add $DEVICE parent 1: protocol ip prio 1 u32 match ip src $SRC2 flowid 1:1 # CBQ qdisc 2:0 $TC qdisc add $DEVICE parent 1:1 handle 2:0 cbq bandwidth $LIMITBW $AVPKT allot 1514 ------------------------------------------------------- -- The statement below is true. The statement above is false. ^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^ Amit Kucheria EECS Grad. Research Assistant University of Kansas @ Lawrence (R)+1-(785)-830 8521 ||| (O)+1-(785)-864 7774 ____________________________________________________
On Thursday 13 December 2001 08:50, you wrote:> On Thu, 13 Dec 2001, bert hubert wrote: > > On Wed, Dec 12, 2001 at 07:17:43PM -0600, Amit Kucheria wrote: > > > I have found CBQ to be very inaccurate (as others on the list) > > > > > > I have created a structure as follow > > > > > > Root (10mbit) > > > > > > CBQ class (bandwidth=1mbit, rate=10mbit) > > > > > > CBQ qdisc (bandwidth=1mbit) > > > > Um, this doesn''t make sense. I keep repeating this to posters here, SHOW > > YOUR CONFIGURATION! Don''t just draw pictures - your CBQ commandlines may > > not do what you think they do. > > My apologies. Here''s the script > > --------------------------------------------------------------------------- > BANDWIDTH=bandwidth 10Mbit > LIMITBW=1Mbit > > # Root CBQ qdisc 1: > $TC qdisc add $DEVICE root handle 1: cbq $BANDWIDTH $AVPKT > > # Root CBQ class 1:1 > # This class rate limits everyting to 1Mbit > $TC class add $DEVICE parent 1:0 classid 1:1 cbq $BANDWIDTH rate $LIMITBW > \ > maxburst 100 $AVPKT allot 1514 weight 100Kbit prio 1 bounded isolated > > # Filter packet from the 2 sources > $TC filter add $DEVICE parent 1: protocol ip prio 1 u32 match ip src > $SRC1 flowid 1:1 > $TC filter add $DEVICE parent 1: protocol ip prio 1 u32 match ip src > $SRC2 flowid 1:1 > > # CBQ qdisc 2:0 > $TC qdisc add $DEVICE parent 1:1 handle 2:0 cbq bandwidth $LIMITBW $AVPKT > allot 1514 > -------------------------------------------------------I''m not seeing any big errors. I tried also the bounded parameter of CBQ, you can find the results on docum.org. I was able to throttle the bandwidth at each speed I want. I din''t attach the second CBQ, but only used the bounded class, maybe you can try again without the second CBQ. And don''t use the isolated parameter. It''s not working and sometimes it can disturb the configuration. Stef -- stef.coene@docum.org More QOS info : http://docum.org/ Title : "Using Linux as bandwidth manager"
On Thu, Dec 13, 2001 at 09:20:32AM +0100, Stef Coene wrote:> > > Um, this doesn''t make sense. I keep repeating this to posters here, SHOW > > > YOUR CONFIGURATION! Don''t just draw pictures - your CBQ commandlines may > > > not do what you think they do. > > > > My apologies. Here''s the scriptAmit, no apologies are needed, it''s just that it helps to get good answers :-)> > # CBQ qdisc 2:0 > > $TC qdisc add $DEVICE parent 1:1 handle 2:0 cbq bandwidth $LIMITBW $AVPKT > > allot 1514Just don''t do this. Having a CBQ within a CBQ is useless and can indeed cause weird behaviour. You will get two qdiscs ''stacked'' so to speak, both trying to shape traffic. At times, the inner qdisc will not want to give up data, which doesn''t make life easier on the outer one. CBQ can do shaping out of its own accord - it doesn''t need a second shaper to do that for it. For the same reasone, people should not embed TBFs in CBQ! Regards, bert -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet
Hello, I have compiled HTB into kernel 2.4.16 and set up simple traffic shaping just to figure how it works. My init script looks as follows: tc qdisc del dev eth1 root tc qdisc add dev eth1 root handle 1: htb default 30 tceth0="tc class add dev eth1" $tceth0 parent 1: classid 1:1 htb rate 128kbit burst 2k $tceth0 parent 1:1 classid 1:10 htb rate 128kbit burst 2k $tceth0 parent 1:1 classid 1:20 htb rate 64kbit ceil 96kbit burst 2k $tceth0 parent 1:1 classid 1:30 htb rate 1kbit ceil 96kbit burst 2k tcqdisc="tc qdisc add dev" $tcqdisc eth1 parent 1:10 handle 10: sfq perturb 10 $tcqdisc eth1 parent 1:20 handle 20: sfq perturb 10 $tcqdisc eth1 parent 1:30 handle 30: sfq perturb 10 u32="tc filter add dev eth1 protocol ip parent 1:0 prio 1 u32" $u32 match ip protocol 1 0xff flowid 1:10 $u32 match ip sport 22 0xffff flowid 1:10 $u32 match ip sport 119 0xffff flowid 1:10 $u32 match ip sport 80 0xffff flowid 1:20 $u32 match ip sport 110 0xffff flowid 1:20 - Pretty much the example script from Advanced Routing HowTo. I have Internet on eth0 and LAN via NAT on eth1. The problem is that after appx. 6 hours or so the local network starts behaving strange - either it doesn''t work at all or just packets from certain classes don''t go through the NAT and get dropped. In syslog I have found following entries: Dec 13 09:33:25 mail kernel: HTB delay 224640 > 5sec Dec 13 09:33:27 mail kernel: HTB delay 224474 > 5sec Dec 13 09:33:29 mail kernel: HTB delay 224272 > 5sec Dec 13 09:33:32 mail kernel: HTB delay 223977 > 5sec Dec 13 09:33:32 mail kernel: HTB delay 223973 > 5sec Dec 13 09:33:37 mail kernel: HTB delay 223473 > 5sec Dec 13 09:33:38 mail kernel: HTB delay 223377 > 5sec Dec 13 09:33:38 mail kernel: HTB delay 223373 > 5sec There is many more of them, they start to appear just at the time when the network starts dying. Any ideas what is wrong pretty please ;) ?? Regards, Leszek
On Thu, Dec 13, 2001 at 10:01:10AM +0100, Leszek Olszewski wrote:> behaving strange - either it doesn''t work at all or just packets from > certain classes don''t go through the NAT and get dropped. In syslog I have > found following entries: > > Dec 13 09:33:25 mail kernel: HTB delay 224640 > 5secTry a newer HTB patch, it fixes this problem. Regards, bert -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet
On Thu, Dec 13, 2001 at 07:48:56AM +0800, yangrunhua wrote:> Thanks for all. Thanks for all the people here and from netfilter > mail-list. > Iproute2+ netfilter seems pretty good! I''ve got good performance! > I''m a new here, so very nice to hear from you, you are so kind.Glad we could help! -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet
Hi Bert,> > Dec 13 09:33:25 mail kernel: HTB delay 224640 > 5sec > > Try a newer HTB patch, it fixes this problem.what is newer? I''m currently experiencing the same problem on a SMP box, kernel 2.4.13 patched with the 39085 bytes htb_2.4.14.diff, HZ set to 1024. tc qdisc add dev eth0 root handle 1: htb default 3 tc class add dev eth0 parent 1: classid 1:1 htb rate 10Mbit burst 10kB tc qdisc add dev eth0 parent 1:1 sfq perturb 10 tc class add dev eth0 parent 1: classid 1:3 htb rate 90Mbit burst 90kB Whenever the overlimit for 1:1 kicks in, "HTB delay > 5sec" messages appear and traffic stalls immediately. Any ideas? Thanks, Stefan
At 17:15 2001-12-14 +0100, Stefan Rompf wrote:>Hi Bert, > > > > Dec 13 09:33:25 mail kernel: HTB delay 224640 > 5sec > > > > Try a newer HTB patch, it fixes this problem. > >what is newer? I''m currently experiencing the same problem on a SMP box, >kernel 2.4.13 patched with the 39085 bytes htb_2.4.14.diff, HZ set to >1024. > >tc qdisc add dev eth0 root handle 1: htb default 3 > >tc class add dev eth0 parent 1: classid 1:1 htb rate 10Mbit burst 10kB >tc qdisc add dev eth0 parent 1:1 sfq perturb 10 >tc class add dev eth0 parent 1: classid 1:3 htb rate 90Mbit burst 90kB > >Whenever the overlimit for 1:1 kicks in, "HTB delay > 5sec" messages >appear and traffic stalls immediately. Any ideas? > >Thanks, StefanYes, I have the same question. I downloaded the most current patch from Devik''s site, patched the kernel, compiled it and again the same problem occurs. Even more, it occurs on a link with virtualy no traffic so there is no overlimit packets at all. Any clue anyone? Regards, Leszek
> > Try a newer HTB patch, it fixes this problem.I haven''t found any newer..> > what is newer? I''m currently experiencing the same problem on a SMPbox,> kernel 2.4.13 patched with the 39085 bytes htb_2.4.14.diff, HZ set to > 1024. > > tc qdisc add dev eth0 root handle 1: htb default 3 > > tc class add dev eth0 parent 1: classid 1:1 htb rate 10Mbit burst 10kB > tc qdisc add dev eth0 parent 1:1 sfq perturb 10 > tc class add dev eth0 parent 1: classid 1:3 htb rate 90Mbit burst 90kB > > Whenever the overlimit for 1:1 kicks in, "HTB delay > 5sec" messages > appear and traffic stalls immediately. Any ideas?I used to have this problem too. I found out that the parent burst was set lower than its children burst. When I increased the parent burst, I didn''t receive more of these messages... I''m not 100% sure if I''ve done the right decision, but it solved the problem. Sincerely Patrik
On Fri, Dec 14, 2001 at 06:24:02PM +0100, Patrik wrote:> > > Try a newer HTB patch, it fixes this problem. > > I haven''t found any newer..Ok - Devik is a bit quiet right now, I think he is busy setting up his internet connection. When he''s back he''ll probably help out. Regards, bert. -- http://www.PowerDNS.com Versatile DNS Software & Services Trilab The Technology People Netherlabs BV / Rent-a-Nerd.nl - Nerd Available - ''SYN! .. SYN|ACK! .. ACK!'' - the mating call of the internet
Yes yes yes .. it all falls at me. I''ve done big bugfix release but because my compiled kernels stopped to work I was nt ablr to test nre patch and thus I didn''t released it yet. For all, there is now hotfix at my page. devik On Fri, 14 Dec 2001, Leszek Olszewski wrote:> At 17:15 2001-12-14 +0100, Stefan Rompf wrote: > >Hi Bert, > > > > > > Dec 13 09:33:25 mail kernel: HTB delay 224640 > 5sec > > > > > > Try a newer HTB patch, it fixes this problem. > > > >what is newer? I''m currently experiencing the same problem on a SMP box, > >kernel 2.4.13 patched with the 39085 bytes htb_2.4.14.diff, HZ set to > >1024. > > > >tc qdisc add dev eth0 root handle 1: htb default 3 > > > >tc class add dev eth0 parent 1: classid 1:1 htb rate 10Mbit burst 10kB > >tc qdisc add dev eth0 parent 1:1 sfq perturb 10 > >tc class add dev eth0 parent 1: classid 1:3 htb rate 90Mbit burst 90kB > > > >Whenever the overlimit for 1:1 kicks in, "HTB delay > 5sec" messages > >appear and traffic stalls immediately. Any ideas? > > > >Thanks, Stefan > > Yes, I have the same question. > > I downloaded the most current patch from Devik''s site, patched the kernel, > compiled it and again the same problem occurs. Even more, it occurs on a > link with virtualy no traffic so there is no overlimit packets at all. > > Any clue anyone? > > Regards, > Leszek > > > > > > _______________________________________________ > LARTC mailing list / LARTC@mailman.ds9a.nl > http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://ds9a.nl/lartc/ > >