This Article 
 Bibliographic References 
 Add to: 
Management of Partially Safe Buffers
March 1995 (vol. 44 no. 3)
pp. 394-407

Abstract—Safe RAM is RAM which has been made as reliable as a disk. We consider the problem of buffer management in partially safe buffers, i.e., buffers which contain both safe RAM and volatile RAM. Buffer management techniques for partially safe buffers explicitly consider the safety of memory in deciding which data to place in the buffer, where to place it, and when to copy updates back to the disk. We present techniques for managing such buffers and study their performance using trace-driven simulations.

[1] M. Baker,S. Asami,E. Deprit,, and J. Ousterhout,“Non-volatile Memory for fast reliable file systems,” Proc. Int’l Conf. Architectural Support Programming Languages and Operating Systems, pp. 10-22, Oct. 1992.
[2] M. Baker and M. Sullivan,“The recovery box: Using fast recovery to provide high availability in the Unix environment,” Proc. Usenix Technical Conf., pp. 31-44, June 1992.
[3] G. Copeland,R. Krishnamurty,, and M. Smith,“The case for safe RAM,” Proc. 15th VLDB Conf., pp. 327-336,Amsterdam, 1989.
[4] M.H. Eich,“A classification and comparison of main memory database recovery techniques,” Proc. Int’l Conf. Data Eng., IEEE, pp. 332-339, Feb. 1987.
[5] J. Gait,“Phoenix: A safe in-memory file system,” Comm. ACM, vol. 33, no. 1, pp. 81-86, Jan. 1990.
[6] H. Garcia-Molina and K. Salem,“High performance transaction processing with memory-resident data,” Proc. Int’l Symp. High Performance Computer Systems,Paris, Dec. 1987,North-Holland, 1988.
[7] T. Haerder and A. Reuter,“Principles of transaction-oriented database recovery,” ACM Computing Surveys, vol. 15, no. 4, pp. 287-317, Dec. 1983.
[8] IBM 3990 Storage Control Introduction, IBM, manual no. GA37-0098-0, 1987.
[9] S.J. Leffler,M.K. McKusick,M.J. Karels,, and J.S. Quarterman,Design and Implementation of the 4.3BSD Unix Operating System,Addison-Wesley, 1989.
[10] T.J. Lehman and M.J. Carey,“A recovery algorithm for a high-performance memory-resident database system,” Proc. ACM SIGMOD Int’l Conf. Management of Data, pp. 104-117,San Francisco, 1987.
[11] S.W. Ng,“Improving disk performance via latency reduction,” IEEE Trans. Computers, vol. 40, no. 1, pp. 22-30, Jan. 1991.
[12] E. Rahm,“Performance evaluation of extended storage architectures for transaction processing,” Proc. ACM SIGMOD Int’l Conf. Management of Data, pp. 308-317, 1992.
[13] C. Ruemmler and J. Wilkes,“Unix disk access patterns,” Usenix Conf. Proc., pp. 405-420, Jan. 1993.
[14] H. Schwetman,“CSIM Reference Manual (Rev. 14),” MCC Technical Report No. ACT-ST-252-87, MCC, Austin, Texas, Mar. 1990.
[15] A.J. Smith, "Disk Cache—Miss Ratio Analysis and Design Considerations," ACM Trans. Computer Systems, Aug. 1985, pp. 161-203.
[16] J.A. Solworth and C.U. Orji,“Write-only disk caches,” Proc. ACM SIGMOD Int’l Conf. Management of Data, pp. 123-132, May 1990.
[17] M. Stonebraker, "The Design of the Postgres Storage System," Proc. VLDB Conf., pp. 289-300, 1987.
[18] M. Stonebraker,R. Katz,D. Patterson,, and J. Ousterhout,“The design of XPRS,” Proc. 14th Int’l Conf. Very Large Data Bases, pp. 318-330, 1988.

Index Terms:
Buffer management, disks, operating systems, performance, reliability, safe RAM.
Kenneth Salem, Sedat Akyürek, "Management of Partially Safe Buffers," IEEE Transactions on Computers, vol. 44, no. 3, pp. 394-407, March 1995, doi:10.1109/12.372032
Usage of this product signifies your acceptance of the Terms of Use.