I've got a conceptual question that I've been wrestling with due to a customer inquiry....I recently had a customer inquiry that stated that he was getting roughly only 1.5 mb transfer speeds over his 2 T-1's to the internet. Excluding other factors such as servers, firewalls internet congestion, etc....should this customer be capable of getting near 3mb download speeds on a individual user session over his 2 Tl's??? Currently, the customer is configured with 2 T1's to the same internet gateway with equal cost static routes on both ends utilizing fast switching with destination load sharing. I'm under the assumption the current set up will only allow for near T1 speeds for a each user session because each transmission will utilize only one T1 with the destination based load sharing causing a user transmission to pick only one path or T1 for the entire session and each T1 is only clocked at 1536mb.
I'm basically wrestling with idea of when a user in this scenario should be capable of getting 3md transfer speeds....
If I use a bundling technique such as MPPP or CEF with per packet load sharing, will that allow a user on this network theoretically to reach near 3mb transfer speeds for a individual user session. Moreover, could someone further explain the differences in terms of latency and transfer speeds from a individual user standpoint with the following network scenarios:
1 - 2 T1's with CEF per packet (equal cost statics)
2 - 2 Tl's with fast switching (equal cost statics/destination load sharing)
3 - 2 T1's imuxed (utlizing added hardware on each end)
thanks for your input,