Confused newb
-
Sorry, i'm sure this is covered in the manual and such somewhere, but i'm not too familiar with all the technical language when it comes to networking, and I wouldn't know what to look for.
First off how low down does the load balancing work? is it at the packet level or would that not work? What I mean is say, you were running a single program with a single connection to a server, like streaming a video, or playing a game online, would that program benefit from the increased speed of the load balancing, or would all that data be routed over only one connection? Also what about usenet (multiple connections to the same server) or bittorrent?
My other question is if this will work in this situation at all, my sister has just moved into student halls, and the provided internet connection is stupidly slow, there is no wireless ISP in range, and although she could get very fast adsl2+ she isn't allowed to have a line connected. and there is no option to pay the university for a faster connection. So while this might seem a little dishonest, there doesn't seem to be any other option.
The connection provided to the halls, is just a regular wired 100mbit ethernet, and as it is split across a couple of hundred rooms and every student has the same speed access, clearly the WAN connection must be a lot faster than the students are getting, and that some sort of limiting/shaping software is being used share the connection. I haven't tested it yet but I don't think there is any contention other than the 100mbit limit.
So I wondered if by building a server with 2 or 3 NICs each with their own IP address (do they have to be static or does DHCP work?) I could load balance the connections and thus get an overall faster connection speed.
Also as far as trying to do things legitimately is concerned, is it likely that the capping is applied per IP address, or MAC address or session ID or what? What I'm wondering is if there is any easy way for the University to be able to charge for a fast connection on a case by case basis, if so a petition for this to be done could be a good idea.
Thanks
-
Sorry, i'm sure this is covered in the manual and such somewhere, but i'm not too familiar with all the technical language when it comes to networking, and I wouldn't know what to look for.
First off how low down does the load balancing work? is it at the packet level or would that not work? What I mean is say, you were running a single program with a single connection to a server, like streaming a video, or playing a game online, would that program benefit from the increased speed of the load balancing, or would all that data be routed over only one connection? Also what about usenet (multiple connections to the same server) or bittorrent?
It's per-connection. You can achieve the aggregate speed when using multiple connections (like with a download manager), but not a single connection (like a HTTP download). That's because of the way TCP/IP works, a single session must be sourced from the same IP, hence it has to stay on one WAN interface.
So I wondered if by building a server with 2 or 3 NICs each with their own IP address (do they have to be static or does DHCP work?) I could load balance the connections and thus get an overall faster connection speed.
Also as far as trying to do things legitimately is concerned, is it likely that the capping is applied per IP address, or MAC address or session ID or what? What I'm wondering is if there is any easy way for the University to be able to charge for a fast connection on a case by case basis, if so a petition for this to be done could be a good idea.
There are countless reasons it could be slow, and countless ways they could be rate limiting traffic. It's possible to use multi-WAN with however many ports you want, static IP or DHCP, but there's no way for us to know if that's going to improve anything. I'd say it's a toss up.