The access-list defines the group of IP addresses, and the rate-limit limit the bandwidth for all the IPs in that ACL (The aggregate), meaning that if you have defined 4 IPs in that ACL, one of the IPs could reach the BW limit if the other don't transmit. I have used rate-limit for such scenarios many times and it worked fine, the only point was defining the BURST SIZE so that the client could reach its maximum limit. If the Busrt Size is not defined well and you create a limit of 1 Mbps, the client might not even reach 900 Kbps.
On my experience, Rate-limit treats the whole ACL and all IPs defined in that ACL as one entity, I don't get what you mean by "the amount of bandwidth specified in the statement will be given on a case-by-case basis." HTH Hamid ""sisco"" wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > buy sitara network box! great graphical bandwidth usage per ip address and > you can even restrict the application ports like kazaa,ftp.... > > > ""s vermill"" wrote in message > [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > > Hamid Ali Asgari wrote: > > > > > > Create an access-list and include all the IP addresses of that > > > group in > > > that access-list. > > > > > > Use rate-limit on the interface to limit the BW for that > > > access-list > > > > > > > Does rate-limiting work like that? I thought that if the condition is met > > (i.e. the address is within the range specified in the ACL), the amount of > > bandwidth specified in the statement will be given on a case-by-case > basis. > > Or does it truly divide the bandwidth amongst all who are allowed by the > ACL? > > > > We once tried to simulate the throughput of a DS3 by creating a policy for > > rate-limiting on a 100 Mbps ethernet. Unfortunately, the machine running > > ttcp to generate the dummy traffic couldn't sustain 45 Mbps. I think they > > ultimately went with two machines, which resulted in more than 45 Mbps of > > traffic but less than 90 Mbps. I seem to recall that the policy ended up > > allowing *each* machine up to 45 Mbps - but I could be wrong. > > Unfortunately, the engineer responsible for that experiment has left for > > greener pastures. > > > > Anyone refresh my memory? Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=53192&t=53066 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
