RE: A Simple Question

["Y P Cheng" <ycheng@advansys.com>]

> From: owner-ips@ece.cmu.edu [mailto:owner-ips@ece.cmu.edu]On Behalf Of
> julian_satran@il.ibm.com
>
> The whole discussion thread is based on the assumption that there will
> be cases in which you will need several adapters (more than one) while you
> keep assuming only one.
>
I guess this highlights all my arguments. :-)

My other two postings on "Asymmetric Model" and "Transport Protocol without
ACK" should give addition background where I come from.

Yes, our adapter is quite capable and will execute hundreds or thousands
iSCSI requests concurrently and atomically, i.e. the adapter will break up
the iSCSI requests into PDU's, order them, and retransmit them.  The adapter
deals with traffic jam, not the device driver.  There is no deadlock problem
in the adapter.  BTW, this is how most fibre channel adapter works today to
deliver FCP requests.  I am not reinventing the wheel.

Yes, it is our intent to support multiple adapters in the device driver for
load balance and recovery.

Yes, it is difficult to synchronizing PDUs of a single iSCSI request in two
or more adapters when they don't share the execution state.