How come some sites which are on dual processor are slower than single processor with same RAM and same HDD.
Even if the response time of dual is better than single processor.
I have come to know something about it. Is it some kind of bandwith trick on the sever.
Like some gives 128 , 256 and 512 connection.
Can anyone tell me something about it.

It could be the connection between you and the server.

Especially over the web and over a dialup link is far more closely linked to performance of the network connection and not the server's configuration.


You can use tools like ping to estimate packet loss and latency between two sites that behave differently to see if there is a difference in the connectivity or effective connectivity.

A dual processor server that is doing 3x the work will probably feel slower all things being equal [compared to a single processor one].

Rik

Maybe those site admins are not properly using their dual processors...

...or they just didn't configure their servers properly in general. Or tune them for the load they will see.

Rik

YES, manveen, you are correct. Your host does LIMIT THE REQUEST. I haven't come across any host who "don't" do that.

Originally posted by oldtimmer
YES, manveen, you are correct. Your host does LIMIT THE REQUEST. I haven't come across any host who "don't" do that.

Thats how some companies offer unlimited bandwith. But is there any way to find out how much bandwith is alloted to the server like 128/256/512.
Some companies gives 300-500GB per month but they limit the request.

Originally posted by manveen


Thats how some companies offer unlimited bandwith. But is there any way to find out how much bandwith is alloted to the server like 128/256/512.
Some companies gives 300-500GB per month but they limit the request.

Yeah, ask ;)

Or run tests yourself. Open up the ftp client and download a file 1MB or greater from each of the following major sites: FTP.UU.NET, FTP.AOL.COM, and FTP.GNU.ORG.

Just one way.

Sometimes, the bandwidth problem isn't as direct as someone "trying" to limit the bandwidth. Sometimes, they are just nitwits when it comes to configuring the switch your server is connected to. For instance, if the switch is set/limited to 10mbps when your server is capable of fast ethernet speeds (100 mbps) or the switch is auto-negotiating to half duplex instead of full duplex.

I agree that ping is a good place to start, but it will only give you the response time. A more informative tool is traceroute, but you need to know how to read it and in these days of "asyncronous" routing, you also have to understand what it isn't telling you. But you should use the traceroute to see where the problem is popping up. Do one from your computer and post the results here. Also, you can try running the traceroute from a third party network like the one available at http://www.valkaryn.net/trace/ or others listed at Traceroute.org. Using multiple traceroute servers can give you a better picture of where the problem is occuring.