Extrapolating from our NFS and iSCSI results, it appears that block- and file-access protocols are comparable on data-intensive benchmarks and the former outperforms the latter on the meta-data intensive benchmarks. From the perspective of performance for IP-networked storage in an unshared environment, this result favors a block-access protocol over a file-access protocol. However, the choice between the two protocols may be governed by other significant considerations not addressed by this work such as ease of administration, availability of mature products, cost, etc.
Observe that the meta-data performance of the NFS protocol suffers primarily because it was designed for sharing of files across clients. Thus, when used in an environment where files are not shared, the protocol pays the penalty of features designed to enable sharing. There are two possible ways to address this limitation: (1) Design a file-access protocol for an unshared environments; and (2) Extend the NFS protocol so that while it provides sharing of files when desired, it does not pay the penalty of ``sharing'' when files are not shared. Since sharing of files is desirable, we propose enhancements to NFS in Section 7 that achieve the latter goal.