Highly scalable web applications with zero-copy data transfer

ABSTRACT

The performance of server-side applications is becoming increasingly important as more applications exploit the Web application model. Extensive work has been done to improve the performance of individual software components such as Web servers and programming language runtimes. This paper describes a novel approach to boost Web application performance by improving inter-process communication between a programming language runtime and Web server runtime. The approach reduces redundant processing for memory copying and the context switch overhead between user space and kernel space by exploiting the zero-copy data transfer methodology, such as the sendfile system call. In order to transparently utilize this optimization feature with existing Web applications, we propose enhancements of the PHP runtime, FastCGI protocol, and Web server. Our proposed approach achieves a 126% performance improvement with micro-benchmarks and a 44% performance improvement for a standard Web benchmark, SPECweb2005.

REFERENCES

• 1: Erich Nahum , Tsipora Barzilai , Dilip D. Kandlur, Performance issues in WWW servers, IEEE/ACM Transactions on Networking (TON), v.10 n.1, p.2-11, February 2002 [doi>10.1109/90.986497]
• 2: Vivek S. Pai , Peter Druschel , Willy Zwaenepoel, IO-lite: a unified I/O buffering and caching system, Proceedings of the third symposium on Operating systems design and implementation, p.15-28, February 1999, New Orleans, Louisiana, United States
• 3: Dragan Stancevic, Zero copy I: user-mode perspective, Linux Journal, v.2003 n.105, p.3, January 2003
• 4: Hsiao-keng Jerry Chu, Zero-copy TCP in Solaris, Proceedings of the 1996 annual conference on USENIX Annual Technical Conference, p.21-21, January 22-26, 1996, San Diego, CA
• 5: Mark R. Brown, "FastCGI: A High-Performance Gateway Interface", Fifth International World Wide Web Conference, 1996
• 6: Mark R. Brown, "Understanding FastCGI Application Performance", http://www.fastcgi.com/
• 7: Arun Iyengar , Jim Challenger, Improving web server performance by caching dynamic data, Proceedings of the USENIX Symposium on Internet Technologies and Systems on USENIX Symposium on Internet Technologies and Systems, p.5-5, December 08-11, 1997, Monterey, California
• 8: Lighttpd, http://www.lighttpd.net/
• 9: PHP, http://www.php.net/
• 10: Philippe Joubert , Robert B. King , Richard Neves , Mark Russinovich , John M. Tracey, High-Performance Memory-Based Web Servers: Kernel and User-Space Performance, Proceedings of the General Track: 2002 USENIX Annual Technical Conference, p.175-187, June 25-30, 2001
• 11: David Pariag , Tim Brecht , Ashif Harji , Peter Buhr , Amol Shukla , David R. Cheriton, Comparing the performance of web server architectures, ACM SIGOPS Operating Systems Review, v.41 n.3, June 2007
• 12: Gaurav Banga, Peter Druschel, Jeffrey C. Mogul, Better operating system features for faster network servers, In Proc. Of the Workshop on Internet Server Performance 1999
• 13: SCGI: A Simple Common Gateway Interface Alternative, http://python.ca/scgi/protocol.txt
• 14: SPECweb2005 http://www.spec.org/web2005/
• 15: Khalil Amiri, Sanghyun Park, et al., "DBProxy: A dynamic data cache for Web applications" ICDE 2003, pages 821--831, 2003
• 16: Zachary Peterson , Randal Burns, Ext3cow: a time-shifting file system for regulatory compliance, ACM Transactions on Storage (TOS), v.1 n.2, p.190-212, May 2005 [doi>10.1145/1063786.1063789]
• 17: Yousef A. Khalidi , Moti N. Thadani, An Efficient Zero-Copy I/O Framework for UNIX, Sun Microsystems, Inc., Mountain View, CA, 1995
• 18: Dong-Jae Kang, Young-Ho Kim, et al., Design and Implementation of Zero-Copy Data Path for Efficient File Transmission, Sep, 2006
• 19: Netcraft, http://survey.netcraft.com/Reports/200806/
• 20: Apache Web Server: http://httpd.apache.org/
• 21: Amol Shukla , Lily Li , Anand Subramanian , Paul A. S. Ward , Tim Brecht, Evaluating the performance of user-space and kernel-space web servers, Proceedings of the 2004 conference of the Centre for Advanced Studies on Collaborative research, p.189-201, October 04-07, 2004, Markham, Ontario, Canada
• 22: Ruby, http://www.ruby-lang.org/en/
• 23: Scott Trent , Michiaki Tatsubori , Toyotaro Suzumura , Akihiko Tozawa , Tamiya Onodera, Performance Comparison of PHP and JSP as Server-Side Scripting Languages, Proceedings of the ACM/IFIP/USENIX 9th International Middleware Conference, December 01-05, 2008, Leuven, Belgium [doi>10.1007/978-3-540-89856-6_9]
• 24: Angelos D. Keromytis , Jason L. Wright , Theo De Raadt , Matthew Burnside, Cryptography as an operating system service: A case study, ACM Transactions on Computer Systems (TOCS), v.24 n.1, p.1-38, February 2006 [doi>10.1145/1124153.1124154]
• 25: Sathish K. Palaniappen, et al., Efficient Data Transfer through zero copy, IBM developerworks, Sep, 2008 http://www.ibm.com/developerworks/library/j-zerocopy
• 26: Francis O'Carroll , Hiroshi Tezuka , Atsushi Hori , Yutaka Ishikawa, The design and implementation of zero copy MPI using commodity hardware with a high performance network, Proceedings of the 12th international conference on Supercomputing, p.243-250, July 1998, Melbourne, Australia [doi>10.1145/277830.277883]
• 27: Nick Mitchell, Gary Sevitsky, Harini, Srivivasan, et al., The Diary of a Datum: An Approach to Modeling Runtime Complexity in Framework-Based Applications , IBM Research Report, RC23703, 2005
• 28: Akihiko Tozawa, Michiaki Tatsubori, Scott Trent, Toyotaro Suzumura, and Tamiya Onodera, P9: High Performance PHP Runtime, Japan Society for Software Science and Technology, 25th Workshop, 2008
• 29: OProfile, http://oprofile.sourceforge.net/
• 30: Netperf, http://www.netperf.org/
• 31: Lakshmish Ramaswamy , Arun Iyengar , Ling Liu , Fred Douglis, Automatic Fragment Detection in Dynamic Web Pages and Its Impact on Caching, IEEE Transactions on Knowledge and Data Engineering, v.17 n.6, p.859-874, June 2005 [doi>10.1109/TKDE.2005.89]
• 32: Michiaki Tatsubori , Toyotaro Suzumura, HTML templates that fly: a template engine approach to automated offloading from server to client, Proceedings of the 18th international conference on World wide web, April 20-24, 2009, Madrid, Spain [doi>10.1145/1526709.1526837]
• 33: Emmanuel Cecchet , Anupam Chanda , Sameh Elnikety , Julie Marguerite , Willy Zwaenepoel, Performance comparison of middleware architectures for generating dynamic web content, Proceedings of the ACM/IFIP/USENIX 2003 International Conference on Middleware, June 16-20, 2003, Rio de Janeiro, Brazil

INDEX TERMS

Primary Classification:
  D. Software
  D.3 PROGRAMMING LANGUAGES
      D.3.4 Processors
          Subjects: Optimization

Additional Classification:
  D. Software
  D.3 PROGRAMMING LANGUAGES
      D.3.4 Processors
          Subjects: Run-time environments

  E. Data
  E.0 GENERAL

  J. Computer Applications
  J.0 GENERAL

General Terms:
Experimentation, Measurement, Performance

Keywords:
PHP, fastcgi, scripting, sendfile, web server, zero copy