Weekly edition	Kernel	Security	Distributions	Contact Us	Search
Archives	Calendar	Subscribe	Write for LWN	LWN.net FAQ	Sponsors

Virtualization and InfiniBand

Posted Aug 8, 2009 7:24 UTC (Sat) by abacus (subscriber, #49001)
In reply to: Virtualization and InfiniBand by dlang
Parent article: AlacrityVM

What you wrote above about cabling is correct but completely irrelevant in this discussion. What I proposed is to use the IB API's (RDMA) and software stack (IPoIB, SDP, iSER, SRP, ...) for communication between a virtual machine and the host system. In such a setup no physical cables are necessary. An additional kernel driver will be necessary in the virtual machine however that implements the RDMA API and allows communication between guest and host.

(Log in to post comments)

Posted Aug 8, 2009 9:27 UTC (Sat) by dlang (✭ supporter ✭, #313) [Link]

if you are talking a virtual interface, why would you use either?

define a driver that does page allocation tricks to move data between the client and the host for zero-copy communication. at that point you beat anything that's designed for a real network.

then you can pick what driver to run on top of this interface, SCSI, IP, custom depending on what you are trying to talk to on the other side.

Virtualization and InfiniBand

Posted Aug 8, 2009 10:52 UTC (Sat) by abacus (subscriber, #49001) [Link]

As I wrote above, implementing an IB driver would allow to reuse a whole software stack (called OFED) and the implementation of several communication protocols. Yes it is possible to develop all this from scratch, but that is more or less like reinventing the wheel.