>>> That's one sloooooow connection - luckily nowadays "the cloud" is a lot
>>> faster, IME companies specialising in that type of engineering
>>> outsourcing work usually have very fast connections (they would have
>>> received a 20MB file before you could call them and say "hello").
>>
>> 200ms from North America to India is extremely fast.
>
> I'm talking about your transfer rate of 20MB in 6 hours being slow. I
> have a 143 ms ping here to www.microsoft.com but can download at over a
> megabyte per second. How are the two linked? If your link took 6 hours
> to download 20MB then it wasn't because of the ping being 200ms.

I suspect that Microsoft will have servers in the UK to send files to 
faster than if you were getting them from Redmond.

Secondly, I never claimed that the network latency was the only reason 
for the length of the transfer, but it is definitely one of the main 
reasons.

This is how it goes (let's assume a well behaved protocol, like FTP):

Every ten packets or so, the sender has to wait for the recipient to 
acknoledge reception of the packets,

Server sends 10 packets...
... it takes 200ms for them to get there.
Client acknowledges...
... it takes 200ms for the acknowlegdement to get back.
Server send 10 more packets...
etc...

Now, as Dr. John said in another e-mail, we have to take into account 
the quality of the Indian telecommunications infrastructure, so let's 
say that 1 packet out of 20 gets dropped, sometimes the scenario becomes:

Server sends 10 packets...
... it takes 200ms for them to get there, but packet #4 got dropped.
Client acknowledges reception of the first 3... packets 5-10 are ignored.
... it takes 200ms for the acknowlegdement to get back.
Server re-sends packets 4-10 and sends 11-13...
etc...

Now, remember that there are supposed to be multiple persons in India 
trying to load different files at the same time, so on top of the 
retransmissions due to poor line quality, you also run the risk of 
having congestions, causing delays in transmission, which will make the 
server assume that none of the packets were received, since they weren't 
acknoleged on time, causing them to be resent, causing further congestion.


-- 
/*Francois Labreque*/#local a=x+y;#local b=x+a;#local c=a+b;#macro P(F//
/*    flabreque    */L)polygon{5,F,F+z,L+z,L,F pigment{rgb 9}}#end union
/*        @        */{P(0,a)P(a,b)P(b,c)P(2*a,2*b)P(2*b,b+c)P(b+c,<2,3>)
/*   gmail.com     */}camera{orthographic location<6,1.25,-6>look_at a }

Post a reply to this message