I'm curious, how does the ISP using a cache server save bandwidt? I can see it if the requests are coming from their clients to web pages outside [..]
Which is the case I'm describing. I, their customer, am seeking a web page outside their network.
but how does it help when web page requests come from outside users to web pages that are inside? The outside bandwidth is still used.
If you just want an academic example of where it can help, think of a Central American ISP. Dedicated ISDN or limited DSL for my house is less than $70, which is a cool deal. Considering the difference in volume of customers between here and the USA, that's downright cheap. But a 100% CIR (carrier-class) link from here to El Salvador is $2000/Mbps, and to Honduras it's around $3500/Mbps (last time I checked). Some "inside" links or parts of your network can benefit greatly from some bandwidth saved.
I'm still confused as to how any kind of DNS lookup is going to make a difference in the routing of the HTTP packets. All DNS is for is to convert the "www.simpaticus.com" into an actual IP address. At that point the box then sends an HTTP request to that IP address.
Which, in most cases, is outside the computer you're using as a client, thus gets sent out, routed, intercepted, and proxied.
In my case, the IP addess is the SAME box that the HTTP request from the browser is coming from and going to. Are you telling me that the Linux routing software will take that IP request and send it to my ISP's router/gateway, only to have it returned to me?
I am not sure whether that happens or not. Regardless, think back for a moment: how is this in any way related to the thread under discussion? I have three Linux servers at home, one of which I use to check my pages while I'm developing. I then upload everything to a test/staging server to make sure it works well on an environment as close as possible to what my webserver has on it. Finally, I rsync the pages to the production server.
All I want to know then, is whether the site actually functions as expected and the pages load as expected on the PRODUCTION machine. There is no way on God's green Earth that anything you've mentioned is going to help me... I think that may be why Jason is losing patience with you. If you're wondering what happens to the routing on your box, then you have a reasonable question although it's off-topic. But if he thinks you're suggesting something _I_ should try, then he's going to think you're out of your mind.
You can test your question, by the way: set up your firewall to block all HTTP requests to your external address, but let Apache listen on that interface. Now, generate a request for a page on that outside IP address. If it gets serviced internally, you'll get a page; if not, you won't.
-- Rodolfo J. Paiz rpaiz@xxxxxxxxxxxxxx http://www.simpaticus.com
-- redhat-list mailing list unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe https://www.redhat.com/mailman/listinfo/redhat-list
