I have seen Duster's web regarding bandwith, I have searched the forum; but I could not find a definite conclusion (not too many discussion discuss this actual method). Maybe Duster can help me out here.

here is one hypothetical case:
my server is only access every half an hour for 5-10 minutes or so... one example bandwith recorded every 30 min.

at 10:00am 128kbps, for 5 min .... transfer: 4.7MB
at 10:30am 512kbps, for 10min ... transfer: 37.5MB
at 11:00am 64kbps, for 10 min ... transfer: 4.7MB
at 11:30am 256kbps, for 5 min ... transfer 9.4 MB

Total actual byte transfer: 56.3MB

Now, average bandwith: (512+128+64+256)/4 = 240kbps

This average will be calculated to produced the average transfer of (During that 2 hours period):
240kbps*60*60*2/(8*1024) = 211 MB

so instead of 56MB, I will be charged by about 211 MB...

would a host calculating bandwith using actual measurement always be cheaper (if not the same/very close) to that using other different method of measurements? If not, can you show me an example where this is not the case? what do you guys think?

cheers,
:beer: