From mboxrd@z Thu Jan 1 00:00:00 1970 From: sagi@grimberg.me (Sagi Grimberg) Date: Mon, 13 Mar 2017 10:32:38 +0200 Subject: [RFC PATCH] nvmet: Back namespace with files In-Reply-To: <20170309174120.GA14329@localhost.localdomain> References: <1489008937-31043-1-git-send-email-keith.busch@intel.com> <20170308213519.GA32009@lst.de> <20170308221527.GA1885@localhost.localdomain> <20170308222014.GA635@lst.de> <20170309174120.GA14329@localhost.localdomain> Message-ID: >> Given that namespace management is tied to controller IDs it's a fairly >> bad fit for fabrics, especially so with our dynamic controller model. > > Hm, implementation details aside for a second, isn't namespace management > more useful on fabrics than pci? It's like managing LUNs on a SAN, > but with spec defined commands. Not exactly, usually namespace/lun provisioning is something that a given host does not typically do or even aware of. Unlike in PCIe, in fabrics, the host does not really own "the" subsystem, it just owns a virtual subsystem that the target exposed for it. If we do support namespace management in the linux target, it'd need to be emulated somehow, obviously a host cannot simply add unprovisioned resources.