Measurements of a distributed file system

Abstract

We analyzed the user-level file access patterns and caching behavior of the Sprite distributed file system. The first part of our analysis repeated a study done in 1985 of the BSD UNIX file system. We found that file throughput has increased by a factor of 20 to an average of 8 Kbytes per second per active user over 10-minute intervals, and that the use of process migration for load sharing increased burst rates by another factor of six. Also, many more very large (multi-megabyte) files are in use today than in 1985. The second part of our analysis measured the behavior of Sprite's main-memory file caches. Client-level caches average about 7 Mbytes in size (about one-quarter to onethird of main memory) and filter out about 50% of the traffic between clients and servers. 35% of the remaining server traffic is caused by paging, even on workstations with large memories. We found that client cache consistency is needed to prevent stale data errors, but that it is not invoked often enough to degrade overall system performance.

Enabling Dynamic Security Management of Networked Systems via Device-Embedded Security (Self-Securing Devices)

“…Second, most read operations complete in a single round trip: reads that observe write concurrency or failures (of storage-nodes or a client write) may incur additional work. Most studies of distributed storage systems (e.g., [Baker9l,Noble94]) indicate that concurrency is uncommon (i.e., writer-writer and writer-reader sharing occurs in well under 1% of operations). Failures, although tolerated, ought to be rare.…”

Increasing Intrusion Tolerance via Scalable Redundancy

“…Second, most read operations complete in a single round trip: reads that observe write concurrency or failures (of storage-nodes or a client write) may incur additional work. Most studies of distributed storage systems (e.g., [Baker91,Noble94]) indicate that concurrency is uncommon (i.e., writer-writer and writer-reader sharing occurs in well under 1% of operations). Failures, although tolerated, ought to be rare.…”

Decentralized Storage Consistency via Versioning Servers

“…In particular, most studies of distributed storage systems (e.g., [4,10,15,26]) indicate that minimal writer-writer and writer-reader sharing occurs, usually well under 1% of operations. In the absence of such sharing, only the uncommon case of partial writes by failing clients requires more than the minimum work for consistency -in a sense, the system is extremely optimistic in that the system only involves extra work if a reader actually observes a concurrent write operation.…”