Hello Michael,
Lemme again thank you for your helpful insight. You have mentioned that
use of tc and iproute from source hasn't given you any issues.
I am gonna following the same instead of using RH packaged tools. Will
let you know the outcome.
Stef....you have any idea about this issue with RH tools???
Trevor
On Fri, 2003-06-27 at 21:01, Michael Ulitskiy wrote:
> On Thursday 26 June 2003 03:22 am, you wrote:
> > Thanks a ton for the same Michael. I am downloading tc from source and
> > will have the same compiled. Am actually using tc as distributed by
> > Martin of HTB fame.
> >
> > So you saying...probably the tc binary distributed by it could be and
> > issue???. Anyway, i will have the same tried out and let you know.
>
> Honestly, I don't know. I'm using iproute2-2.4.7-now-ss020116-try.tar.gz and
> htb3.6-020525.tgz patch and it seems to work fine.
> It's already in production and looks good so far.
>
> Michael
>
> > Trevor
> >
> >
> > On Thu, 2003-06-26 at 01:48, Michael Ulitskiy wrote:
> > > Hi,
> > >
> > > This is the script I'm using to create tc configuration to shape
> > > currently about 8000 ips separately so that each ip has guaranteed
> > > bandwith and cannot exceed it. The script can be used to
> > > shape up to class B. The configuration is using double hashing
> > > for filtering rules. I haven't used it in production yet, but it shows
> > > pretty good result on a testbed.
> > > To test it further I've created configuration to shape range
> > > 192.168.1.0-192.168.254.0 (almost class B), i.e.about 65000 classes
> > > and about 65000 filters. It takes very long time to apply on my
> > > rather slow (P166mmx) test machine, but it doesn't throw a word
> > > of complaint.
> > >
> > > Creating configuration:
> > > root@xxxx:~/perl/qos# time ./qos.conf
> > >
> > > real 0m42.733s
> > > user 0m40.950s
> > > sys 0m0.490s
> > >
> > > It creates file rc.qos. Applying configuration:
> > > root@xxxx:~/perl/qos# time . ./rc.qos
> > >
> > > real 125m48.033s
> > > user 28m40.660s
> > > sys 95m59.420s
> > >
> > > root@xxxxxxx:~/perl/qos# tc qdisc ls | wc -l
> > > 64520
> > > root@xxxxxxx:~/perl/qos# tc filter ls dev eth0 parent 13:0 | wc -l
> > > 129042
> > > (2 output lines for a filter).
> > >
> > > So the answer to your question most likely is no. There's no limit.
> > > Hint: I've experienced something like that when I tried to use precompiled
> > > binary (actually I compiled it myself but on different machine).
> > > The errors were gone after I recompiled tc from the same sources
> > > on the machine where it's being used. Could be some libraries
> > > incompatibilities or changes made to some header files.
> > > So look at the script if you're interested.
> > >
> > > Michael
> > >
> > > p.s. I'm not sure that larts list manager will allow attachment through.
> > > If it doesn't and you want the script - email me.
> > >
> > >
> > > On Tuesday 24 June 2003 06:13 pm, Trevor Warren wrote:
> > > > Hello,
> > > >
> > > > If this isn't a limit of U32 filters then whats wrong with the way i
> > > > have set things up???. The issue has to be somewhere right.
> > > >
> > > > Also, i do appreciate you not wanting to give me your config. But at
> > > > least give me a couple of lines of your U32 filter usage such that i can
> > > > make 5000> filters using it and test. Only passing your u32 filter
> > > > config.....if its your equivalent to IP then its fine with me...:)
> > > >
> > > > I will have the test script mailed across to all. I am in DESPERATE
> > > > need for help on this part......:(
> > > >
> > > > Trevor
> > > >
> > > >
> > > > On Wed, 2003-06-25 at 12:55, Catalin BOIE wrote:
> > > > > > Hello,
> > > > > Hello!
> > > > >
> > > > > > I do understand that limits on u32 filters > 2000 have never been
> > > > > > encountered. Run this script which is base on htb and lemme know. You
> > > > > > will need 2.4.20 > and tc compiled for htb.
> > > > >
> > > > > It gives me a lot of "RTNETLINK answers: File exists".
> > > > > This doesn't seems to be a limit.
> > > > >
> > > > > > Can you please send me your config and lemme compare the same against
> > > > > > mine.
> > > > > I have main in a database. I cannot give it to you, but trust me that are
> > > > > 50000 filters.
> > > > >
> > > > > Try to use hashes to reduce the kernel pressure.
> > > > >
> > > > > >
> > > > > > Trevor
> > > > > >
> > > > > >
> > > > > > On Tue, 2003-06-24 at 15:49, Catalin BOIE wrote:
> > > > > > > On Tue, 24 Jun 2003, Trevor Warren wrote:
> > > > > > >
> > > > > > > >
> > > > > > > > Can some one please confirm which would be the best place to clarify
> > > > > > > > these U32 issues and 2048 U32 filter limitations as i had posted in my
> > > > > > > > previous mail.
> > > > > > > >
> > > > > > > > Trevor
> > > > > > >
> > > > > > > We have 50000 filters and still working.
> > > > > > > I don't think there is a limit.
> > --
> > ( >- LINUX, It's all about CHOICE -< )
> > /~\ __ twarren@xxxxxxxxxx __ /~\
> > | \) / Pre Sales Consultant - Red Hat \ (/ |
> > |_|_ \ 9820349221(M) | 22881326(O) / _|_|
> > \___________________________________/
> >
> >
--
( >- LINUX, It's all about CHOICE -< )
/~\ __ twarren@xxxxxxxxxx __ /~\
| \) / Pre Sales Consultant - Red Hat \ (/ |
|_|_ \ 9820349221(M) | 22881326(O) / _|_|
\___________________________________/
