Hi all 803 list members! This is your list maintainer & HOWTO author speaking. I''ve recently noted a lot of people leaving the list (although a lot of people have also been joining) - is anything wrong? I don''t actually follow the list at all times, please let me know of any problems. Since posting has been restricted to members only, a lot of spam has been blocked which is good. However, a lot of bonafide mail has also been blocked because people aren''t posting from the address they subscribed from. And I know for a fact that I lost one modification request for somebody that moved. If this is you, please resend. In other news, I added a section on HTB, http://ds9a.nl/2.4Routing/HOWTO//cvs/2.4routing/output/2.4routing-10.html#ss10.2 which you should read if you thing CBQ is complicated. And I added a link to docum.org, which you should visit. You should also visit http://www.linux-kongress.de - As you can see from the program http://www.linux-kongress.de/program.html a lot of cool people will be there. And I''ll be there too, and even give a lecture. Hope to meet you there! 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
Jason Tackaberry
2001-Nov-11 21:07 UTC
Re: Everybody happy with the list + some new material
Hi Bert, On Sun, 2001-11-11 at 15:36, bert hubert wrote:> In other news, I added a section on HTB, > http://ds9a.nl/2.4Routing/HOWTO//cvs/2.4routing/output/2.4routing-10.html#ss10.2 > which you should read if you thing CBQ is complicated. And I added a link to > docum.org, which you should visit.According to the HTB patch, HTB stands for Hierarchical Token Bucket. I guess the impression I get from what you wrote is that HTB is good in simple cases, and CBQ is good in complex cases? It might be worth explaining a situation in which HTB is inadequate but CBQ would do the job. I''m not totally clear on that. :) Cheers, Jason.
On Sun, Nov 11, 2001 at 04:07:04PM -0500, Jason Tackaberry wrote:> Hi Bert, > > On Sun, 2001-11-11 at 15:36, bert hubert wrote: > > In other news, I added a section on HTB, > > http://ds9a.nl/2.4Routing/HOWTO//cvs/2.4routing/output/2.4routing-10.html#ss10.2 > > which you should read if you thing CBQ is complicated. And I added a link to > > docum.org, which you should visit. > > According to the HTB patch, HTB stands for Hierarchical Token Bucket.Ah! I looked a bit, but not too well it appears. Will update the HOWTO.> I guess the impression I get from what you wrote is that HTB is good in > simple cases, and CBQ is good in complex cases? It might be worth > explaining a situation in which HTB is inadequate but CBQ would do the > job. I''m not totally clear on that. :)Hmm, well, I''m trying very hard to understand CBQ because I need to give a lecture about it. And I am not succeeding. HTB currently has some performance design issues, which you can read on the site I mention in the HOWTO, but these appear to be resolveable. Right now, if your link sharing needs can be handled by CBQ.init (also mentioned in the HOWTO), by all means use it. I suspect CBQ may scale better to larger amounts of bandwidths and complexer situations. However, if what you want is too difficult for CBQ.init and you have to type the scripts by hand, HTB wins -hands down-. HTB will get there (it is not yet in the kernel for example), but may have some problems today - although authors tend to overestimate the problems their inefficiencies will cause. Right now HTB is polled too often to see if it has a packet available. But just try it! http://luxik.cdi.cz/~devik/qos/htb/ The patch is very modular and does not touch other parts of the kernel. In fact, it could be distributed as a separate program, you just need to have the kernel headers available. 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
> Hmm, well, I''m trying very hard to understand CBQ because I need to give a > lecture about it. And I am not succeeding. HTB currently has someYou don''t know how relieved I am to hear you say that. :) I too am having problems getting a solid handle on CBQ. And here I thought I was just being dense. :)> Right now, if your link sharing needs can be handled by CBQ.init (also > mentioned in the HOWTO), by all means use it. I suspect CBQ may scale betterCBQ.init looks good, and might do the job, but I don''t want to use it on principle because it''s basically a black box. I tried to read and understand the script, but that''s difficult because I don''t understand CBQ. So if I can do everything "from first principles" with HTB, I''d be much happier. I do plan to try it next week. I''ll share my results with everyone. Regards, Jason.
Hi Bert, by the way, what is lartc@mailman.ds9a.nl ?> > > In other news, I added a section on HTB, > > > http://ds9a.nl/2.4Routing/HOWTO//cvs/2.4routing/output/2.4routing-10.html#ss10.2 > > > which you should read if you thing CBQ is complicated. And I added a link to > > > docum.org, which you should visit.thanks for good work and for mentioning HTB :)> > I guess the impression I get from what you wrote is that HTB is good in > > simple cases, and CBQ is good in complex cases? It might be worth > > explaining a situation in which HTB is inadequate but CBQ would do the > > job. I''m not totally clear on that. :)Well, there are: HTB don''t allow you to attach classifiers (filters) to inner levels of hierarchy yet. I''ll implement it if people will want it. This is no deficiency about it only more classifiers per single qdisc might be simpler to maintain. CBQ allows you to play with overlimit strategies (when class is over what to do: delay class, drop class, delay parent ..). HTB always delays class (like TBF or CBQ''s default mode). Almost everyone uses this mode. Maybe some shiny day I''ll add drop mode - it is only mode (except class delay) which makes sense and could shorten delays. And finally, HTB has less knobs to play with. It''s good for majority of people but there are men which like to play with many settings :-) About performance issues see bellow.> Hmm, well, I''m trying very hard to understand CBQ because I need to give a > lecture about it. And I am not succeeding.Drop me a mail. I''ll explain you those dark parts of CBQ. It is not too hard but is it hard enough to make you headaches ;-) I spent year to understand it. But I was still missing something. The found the key to understand it while designing HTB. If you want to understand CBQ then don''t think about top level sharing. It is only way how to implement principle not principle itself...> HTB currently has some > performance design issues, which you can read on the site I mention in the > HOWTO, but these appear to be resolveable. > > Right now, if your link sharing needs can be handled by CBQ.init (also > mentioned in the HOWTO), by all means use it. I suspect CBQ may scale better > to larger amounts of bandwidths and complexer situations. However, if what > you want is too difficult for CBQ.init and you have to type the scripts by > hand, HTB wins -hands down-. > > HTB will get there (it is not yet in the kernel for example), but may have > some problems today - although authors tend to overestimate the problems > their inefficiencies will cause. Right now HTB is polled too often to see if > it has a packet available.Let me correct it a bit. The problem with polling too often is not tied to HTB. All other non-work-conserving disciplines (those which can "hold" packets - HTB,CBQ and TBF) suffer from it. There is no way to say QoS layer: hey here is skb (packet) and it is LAST one I have for now. Instead after each last packet there is one more dequeue event to see whether qdisc is empty. But non-work-conserving qdisc can hold packet and what is worse the packet can be hidden in their non-work-conserving child. Then the qdisc has to do whole work because it doesn''t know whether the child would like to release its skb. As I''ve written in manual, I plan to add option to say that the HTB instance has only work-conserving children (normal use, it sometimes doesn''t hold for experiments and simulations) and it would enable to optimize for such case (quickly return from dequeue when time shows that I just dequeued last packet). So that only performance difference is from different algorithm. Sometimes wins HTB sometimes CBQ. But they are not too different. I did measurements (using pentium''s TSC) and it showed that neither CBQ nor HTB is bottleneck. Most of time is lost somewhere in networking code (80% for 7 class CBQ/HTB) not in QoS. In fact during testing I successfully reached 380 Mbit throughtput with PII/600 and HTB with 7 classes (not tested with CBQ). It was with two full-duplex RTL NICs. I''d sometime like to test gigabit cards ..> The patch is very modular and does not touch other parts of the kernel. In > fact, it could be distributed as a separate program, you just need to have > the kernel headers available.It is true. I''d only note that I fixed the patches a few days ago because they didn''t work as non-module. There was missing initialization call from sch_api.c . But if you want to compile only with includes (as mentioned above by bert) you can simply remove the last ten lines from new patches. best regards, devik
Hi Bert - And thanks for asking, I appreciate it. I''ll offer some feedback. I have a problem with routing that is driving me crazy. I am not able to make a Linux 2.4 kernel (Red Hat V7.1) route ICMP Echo REPLY packets back to the original PINGer. After studying the HOWTO and finding this list, I spent a bunch of time writing up the problem, drew up a really bad ASCII art picture and submitted the problem. It turned out, the list wraps the text, so that totally messed up the ASCII part picture I drew. But then nobody even replied telling me my ASCII art was bad. In fact, nobody even replied at all, other than the automated replies from the list system. And that goes to a core issue for me - I really want to use these nifty advanced routing capabilities with my customers, and I am willing to experiment and learn, but I need to know there really is a community out there to help with issues before I trust this stuff. I know everyone who posts to this list is volunteering time and not getting paid, and I really appreciate any help I can get on my issues. And since I asked for help, I also tried to help a couple other folks as best as I knew how. After all, fair is fair. But if I can''t get answers to my basic questions here, I need to find answers somewhere else. thanks - Greg Scott -----Original Message----- From: bert hubert [mailto:ahu@ds9a.nl] Sent: Sunday, November 11, 2001 2:36 PM To: lartc@mailman.ds9a.nl Subject: [LARTC] Everybody happy with the list + some new material Hi all 803 list members! This is your list maintainer & HOWTO author speaking. I''ve recently noted a lot of people leaving the list (although a lot of people have also been joining) - is anything wrong? I don''t actually follow the list at all times, please let me know of any problems. Since posting has been restricted to members only, a lot of spam has been blocked which is good. However, a lot of bonafide mail has also been blocked because people aren''t posting from the address they subscribed from. And I know for a fact that I lost one modification request for somebody that moved. If this is you, please resend. In other news, I added a section on HTB, http://ds9a.nl/2.4Routing/HOWTO//cvs/2.4routing/output/2.4routing-10.html#ss 10.2 which you should read if you thing CBQ is complicated. And I added a link to docum.org, which you should visit. You should also visit http://www.linux-kongress.de - As you can see from the program http://www.linux-kongress.de/program.html a lot of cool people will be there. And I''ll be there too, and even give a lecture. Hope to meet you there! 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/2.4Routing/
Anton V. Gnitko
2001-Nov-12 15:36 UTC
Re: Everybody happy with the list + some new material
Yea... I would be interesting to know that too... Probably anyone do. On Monday 12 November 2001 12:23, you wrote:> Hi Bert - > > And thanks for asking, I appreciate it. I''ll offer some feedback. I have > a problem with routing that is driving me crazy. I am not able to make a > Linux 2.4 kernel (Red Hat V7.1) route ICMP Echo REPLY packets back to the > original PINGer. After studying the HOWTO and finding this list, I spent > a bunch of time writing up the problem, drew up a really bad ASCII art > picture and submitted the problem. It turned out, the list wraps the text, > so that totally messed up the ASCII part picture I drew. > > But then nobody even replied telling me my ASCII art was bad. In fact, > nobody even replied at all, other than the automated replies from the list > system. > > And that goes to a core issue for me - I really want to use these nifty > advanced routing capabilities with my customers, and I am willing to > experiment and learn, but I need to know there really is a community out > there to help with issues before I trust this stuff. > > I know everyone who posts to this list is volunteering time and not getting > paid, and I really appreciate any help I can get on my issues. And since I > asked for help, I also tried to help a couple other folks as best as I knew > how. After all, fair is fair. But if I can''t get answers to my basic > questions here, I need to find answers somewhere else. > > thanks > > - Greg Scott > > > -----Original Message----- > From: bert hubert [mailto:ahu@ds9a.nl] > Sent: Sunday, November 11, 2001 2:36 PM > To: lartc@mailman.ds9a.nl > Subject: [LARTC] Everybody happy with the list + some new material > > > Hi all 803 list members! > > This is your list maintainer & HOWTO author speaking. I''ve recently noted a > lot of people leaving the list (although a lot of people have also been > joining) - is anything wrong? I don''t actually follow the list at all > times, please let me know of any problems. > > Since posting has been restricted to members only, a lot of spam has been > blocked which is good. However, a lot of bonafide mail has also been > blocked because people aren''t posting from the address they subscribed > from. And I know for a fact that I lost one modification request for > somebody that moved. If this is you, please resend. > > In other news, I added a section on HTB, > http://ds9a.nl/2.4Routing/HOWTO//cvs/2.4routing/output/2.4routing-10.html#s >s 10.2 > which you should read if you thing CBQ is complicated. And I added a link > to docum.org, which you should visit. > > You should also visit http://www.linux-kongress.de - As you can see from > the program http://www.linux-kongress.de/program.html a lot of cool people > will be there. And I''ll be there too, and even give a lecture. > > Hope to meet you there! > > Regards, > > bert-- Sincerely, Anton V. Gnitko ---------------------------------------------------------- Internet Project Manager Central Asian Development Agency, Inc. (CADA) Dushanbe, Tajikistan Tel: (+992 372) 217159 Alt Tel: (+992 372) 212210 Fax: (+992 372) 510052 WWW: http://www.tajik.net
Sebastian ''spax'' Pape
2001-Nov-12 18:36 UTC
Re: Everybody happy with the list + some new material
hi Bert,
I''m quite happy with this list - the only disturbing thing that is
annoying me, is the "quotation-style" here. There are lot''s
of
fullquotes even if most of them are unneccesarry, because the author
only says somthing like: "add me, too". It costs lot of time finding
the essential two or three sentences and it''s lot of unneccesarry
traffic, too. If someone really likes to read all previous mails of
the regarding thread it is possible to look at the web-archive. I
would appreciate it, if people would use only relevant quotes.
best regards
Sebastian
--
Sebastian ''spax'' Pape | Confidence is the feeling you
have before
mailto: sebastian@p-a-p-e.de | you understand the situation.
pgp: http://p-a-p-e.de/pgp.asc |
--- Do you want to know more? http://www.p-a-p-e.de/ ---
On Mon, Nov 12, 2001 at 07:36:13PM +0100, Sebastian ''spax'' Pape wrote:> hi Bert, > > I''m quite happy with this list - the only disturbing thing that is > annoying me, is the "quotation-style" here. There are lot''s of > fullquotes even if most of them are unneccesarry, because the author > only says somthing like: "add me, too". It costs lot of time findingOk, after reading some of the responses to my original question, I''ve added this to the ''new subscriber'' message: Hi! This list is about asking questions and answering them. Please keep that in mind - in true community style, please try to participate! Also, if at all possible, try to trim the amount of text you quote when replying - just leaving the entire message there does not enhance clarity and it also wastes space and bandwidth! Perhaps this can address your issues. Thanks for responding! 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