ACM Home Page
Please provide us with feedback. Feedback
The design and implementation of a log-structured file system
Full text PdfPdf (1.97 MB)
Source ACM Transactions on Computer Systems (TOCS) archive
Volume 10 ,  Issue 1  (February 1992) table of contents
Pages: 26 - 52  
Year of Publication: 1992
ISSN:0734-2071
Authors
Mendel Rosenblum  Univ. of California, Berkeley
John K. Ousterhout  Univ. of California, Berkeley
Publisher
ACM  New York, NY, USA
Bibliometrics
Downloads (6 Weeks): 85,   Downloads (12 Months): 619,   Citation Count: 195
Additional Information:

abstract   references   cited by   index terms   review   collaborative colleagues  

Tools and Actions: Request Permissions Request Permissions    Review this Article  
DOI Bookmark: Use this link to bookmark this Article: http://doi.acm.org/10.1145/146941.146943
What is a DOI?

ABSTRACT

This paper presents a new technique for disk storage management called a log-structured file system. A log-structured file system writes all modifications to disk sequentially in a log-like structure, thereby speeding up both file writing and crash recovery. The log is the only structure on disk; it contains indexing information so that files can be read back from the log efficiently. In order to maintain large free areas on disk for fast writing, we divide the log intosegmentsand use a segment cleaner to compress the live information from heavily fragmented segments. We present a series of simulations that demonstrate the efficiency of a simple cleaning policy based on cost and benefit. We have implemented a prototype log-structured file system called Sprite LFS; it outperforms current Unix file systems by an order of magnitude for small-file writes while matching or exceeding Unix performance for reads and large writes. Even when the overhead for cleaning is included, Sprite LFS can use 70% of the disk bandwidth for writing, whereas Unix file systems typically can use only 5–10%.


REFERENCES

Note: OCR errors may be found in this Reference List extracted from the full text article. ACM has opted to expose the complete List rather than only correct and linked references.

 
1
BAKER, H. G. List processing in real time on a serial computer. A.I. Working Paper 139, MIT-AI Lab, Boston, MA, Apr. 1977.
2
 
3
4
5
 
6
 
7
8
 
9
KAZAR, M. L, LEVERETT, B. W., ANDERSON, O. T, APOSTOLIDES, V., BOTTOS, B. A CHUTANI, S., EVERHART, C. F., MASON, W. A., Tu, S. T., AND ZAYAS, E. R. Decorum file system architectural overview. In Proceedings of the USENIX 1990 Summer Conference (Anaheim, Calif., Jun. 1990), pp. 151-164.
10
11
12
 
13
McKusIcK, M K., JoY, W. N., LEFFLER~ S. J., AND FABRY, R. S. Fsck--The UNIX file system check program. Unix System Manager's Manual--4.3 BSD Virtual VAX-11 Version, USENIX, Berkeley, Calif., Apr. 1986.
 
14
McVoY, L. AND KL~mAN, S. Extent-like performance from a UNIX file system. In Proceedings of the USENIX 1991 Winter Conference (Dallas, Tex., Jan. 1991).
15
 
16
OUSTERHOU% J. K. Why aren't operating systems getting faster as fast as hardware? In Proceedings of the USENIX 1990 Summer Conference (Anaheim, Calif., Jun. 1990), pp. 247-256.
 
17
18
19
 
20
REED, D. AND SVOBODOVA, L. SWALLOW: A distributed data storage system for a local network. In Local Networks for Computer Communications. North-Holland, 1981, pp. 355-373.
 
21
SANDSERG, R. Design and implementation of the Sun network filesystem. In Proceedings of the USENIX 1985 Summer Conference (Portland, Ore , Jun. 1985) pp. 119-130.
 
22
SALEM, K. AND GARCIA-MOLINA, H. Crash recovery mechamsms for main storage database systems. CS-TR-034-86, Princeton Univ., Princeton, NJ 1986.
23
 
24
SELTZER, M. I., CHEN, P. M., AND OUSTERHOUT, J. K Disk scheduling revisited. In Proceedlngs of the Winter 1990 USENIX Techmcal Conference (Jan. 1990), pp.

CITED BY  195


REVIEW

"Charles N. Schroeder : Reviewer"

This well-written paper is of particular interest at this time since it deals with a much-improved disk storage management system for UNIX-based systems. The system described shows an order of magnitude increase in performance over  more...

Collaborative Colleagues:
Mendel Rosenblum: colleagues
John K. Ousterhout: colleagues