[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: NodeOS questions
>1) Flow management
>
>Our question involves the implementation of more complex protocols,
>such as reliable transport, in the EE. In this case, a flow of packets
>for a given AA will contain some packets that will be processed by the
>AA and others that will only be processed by the EE. The later includes
>packets such as SYN, ACKS, and retransmissions.
Perhaps this is too far off the mark, but couldn't the AA be written
to make calls to the EEs toolbox of functions for processing SYNs,
ACKs and retransmission? The AA implementor might be provided with a
base class that makes the right EE calls, and would then implement a
subclass with their application-specific behavior added.
This isn't an answer to your question -- I don't see any obvious ways
of handing things around the flow hierarchy in the current nodeOS
interface spec -- the MAP_SHARED attribute proposed for mempools is
not currently in the spec. I'll let others speak to the thread issue.
--Steve
I'd view this differently. Think of the EE as a library that processes
messages on their way to the AA. You'd set up a sub-flow, with its own
channel, to catch all packets (both control and data) that support the
AA's communication. The main EE flow may have its own channel that SYN
packets arrive on, but this is a separate issue.
Larry
[ Janos ] [ OSKit ] [ Network Testbed ] [ Flick ] [ Fluke ]
Flux Research Group / Department of Computer Science / University of Utah