You have totally misunderstood what I was saying, and it was in my original email I hope no one else took it this way and just think I am a total idiot. My company where the vpn concentrators are located has 2 ds3's to the internet, that is 45mb on each ds3 with a total 90mb to the net. What we used for testing the client piece was a dsl connection and a cable connection and dialup I hope this clears things up for you. I do not know where you are located but even in this day and age bandwidth is still extremely expensive and a 100m conection wow you make it seem as if everyone has this type of connection where you are. And I would rather have 2 45mb connections than one 100mb anyday, reason redundancy and a tad bit of load balancing tho not true load balancing with bgp. We get full routes from both providers which are 2 different tier1 providers Sprint and UUnet and we distribute the default route from both into our IGP. This is great for when you have a farmer doing some digging and happens to cut through UUnet's pipe which happens to be carrying about 128 ds'3 that can't get put back together until about 7 hours later. :) Sorry for the confusion
> That is amazing. I would never have constructed a VPN infrastructure > with 3000 users and 128k. > > In my office, we have a 100m pipe and this is not unusual given the > area. When I was working at Worldcom, we were designing a massive > Metro network infrastructure upgrade for the financial district of > Tokyo. That was back in 2000. I left Worldcom, before the burst mind > you, and now in this same area, most companies have 100m connections > to the Internet. Home DSL is 30m or in my area only 8m with a > dependable 3.5m connection. > > Not bad for living next to rice paddies I think. > > If I had 3000 users, I would have recommended a 1.5 line to the > Internet at least. I was troubleshooting a VPN 5008 problem last year > with 1000 users between New Zealand, Singapore, and Tokyo and that > company had upgraded from 128 to 1.5 and said the performance was just > so much better. > That was for an office of only 50 people local to Tokyo. > > Thanks for your info. I will try to implement it in my designs. > > Theodore Stout, CISSP > Senior Security Consultant > CCNP, CCDP, CSS1, CCSE > > > > > > > "Elijah Savage" > 06/04/2002 08:02 PM > Please respond to esavage > > > To: > cc: , , > Subject: Re: VPN Overhead > [7:45719] > > > The bandwidth of the dsl that the telco generously loaned us for 60 > days was 608 down 128 up, that is another thing if you have a good > working relation with your local telco ask them to get you a circuit > for testing and they will probably do it. >> What is the bandwidth of you DSL? >> >> >> >> >> >> >> >> "Elijah Savage" >> Sent by: [EMAIL PROTECTED] >> 06/04/2002 11:47 AM >> Please respond to "Elijah Savage" >> >> >> To: [EMAIL PROTECTED] >> cc: >> Subject: Re: VPN Overhead [7:45719] >> >> >> We have 2 3030 concentrators setup in a load balancing fashion and it >> works very well. We have rolled this out to about 3000 users and have >> done all types of testing with different applications and different >> types of access. Over dialup we notice that there is about 12% >> overhead with the cisco vpn client, with broadband it makes less of an >> impact. We noticed on broadband that it was about 7% on dsl and about >> 5% on cable access. Hope that helps out. >>> We are currently using a VPN provider to get into the network but >>> want to take more control and bring it in house. I did some testing >>> though and found that the VPN was adding about 27% overhead compared >>> to bypassing VPN and going direct to a server. >>> >>> I'm wondering if others have done testing and what were your results. >>> We are currently using V-One but I will be looking at Cisco's >>> solution. Nondisclosure violations to [EMAIL PROTECTED] >> Nondisclosure violations to [EMAIL PROTECTED] Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=45821&t=45719 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
