Existing metadata-private messaging systems are either non-scalable or vulnerable to long-term traffic analysis. Approaches that mitigate traffic analysis attacks often suffer from unrealistic and unimplementable assumptions or impose system-wide bandwidth restrictions, degrading usability, and performance. In this work, we present a new model for metadata-private communication systems-deferred retrieval-that guarantees traffic analysis resistance under realistic, implementable user assumptions. We introduce Sparta systems, practical and scalable instantiations of deferred retrieval that are distributable, achieve high throughput, and support multiple concurrent conversations without message loss. Specifically, we present three Sparta constructions optimized for different scenarios: (i) low-latency, (ii) high-throughput in shared-memory environments (multi-thread implementations), and (iii) high throughput in shared-nothing (distributed) environments. Our low latency Sparta supports latencies of less than one millisecond, while our high-throughput Sparta can scale to deliver over 700,000 100B messages per second on a single 48-core server.

We have witnessed a 50 million-fold increase in hard disk drive density without a similar increase in performance. How can this unbalanced growth be possible? Can it continue? Can similar unbalanced growth happen in other media? To answer these questions we contrast the value of information storage services with the value of physical storage services. We describe a methodology that separates the costs of capturing, storing and accessing information, and we will show that these aspects of storage systems are independent of each other. We provide arguments for what can happen if the cost of storage continues to decrease. The conclusions are three-fold. First, as capacity of any storage media grows, there is no inherent requirement that performance increase at the same rate. Second, the value of increased capacity devices can be quantified. Third, as the cost of storing information approaches zero, the quantity of information stored will grow without limit.

No abstract available.

Shingled Magnetic Recording (SMR) disks employ a shingled write process that overlaps the data tracks on the disk surface like the shingles on a roof, thereby increasing disk areal density with minimal manufacturing changes. While these disks have the same read behavior as current disks, random writes and in-place data updates are no longer possible, since a write to a track must overwrite and destroy data on all tracks that it overlaps.

Checksum and cyclic redundancy check (CRC) algorithms have historically been studied under the assumption that the data fed to the algorithms was uniformly distributed. This paper examines the behavior of checksums and CRCs over real data from various UNIX file systems. We show that, when given real data in small to modest pieces (e.g., 48 bytes), all the checksum algorithms have skewed distributions. These results have implications for CRCs and checksums when applied to real data. They also can cause a spectacular failure rate for both the TCP and ones-complement Fletcher (1983) checksums when trying to detect certain types of packet splices. When measured over several large file systems, the 16 bit TCP checksum performed about as well as a 10-bit CRC. We show that for fragmentation-and-reassembly error models, the checksum contribution of each fragment are, in effect, colored by the fragment's offset in the splice. This coloring explains the performance of Fletcher's sum on nonuniform data, and shows that placing checksum fields in a packet trailer is theoretically no worse than a header checksum field. In practice, the TCP trailer sums outperform even Fletcher header sums.

Checksum and CRC algorithms have historically been studied under the assumption that the data fed to the algorithms was entirely random. This paper examines the behavior of checksums and CRCs over real data from various UNIX® file systems. We show that, when given real data in small to modest pieces (e.g., 48 bytes), all the checksum algorithms have skewed distributions. In one dramatic case, 0.01% of the check values appeared nearly 19% of the time. These results have implications for CRCs and checksums when applied to real data. They also cause a spectacular failure rate for the both TCP and Fletcher's checksums when trying to detect certain types of packet splices.

The HIPPI data channel and how it is used from a standard data channel, a standard internetworking, and a standard protocol viewpoint are described. Performance is also briefly discussed, as are some existing implementations of switches and other HIPPI realizations.>