Speeding up Database Access with mod_perl

Continuing the discussion of mod_perl, Mr. Lerner tells us about the DBI specification and the Apache::DBI module.
How Fast is it?

Now that we have our test program, how can we actually test it? I wrote a short Perl program using Benchmark.pm, in Listing 3. The timethese function is imported by Benchmark.pm, which we bring in at the beginning of the program. We also bring in LWP::Simple, part of the “Library for WWW access in Perl” that makes it a snap to write a small web client. How simple? Well, the following one-liner returns the HTML contents at http://www.linuxjournal.com/:

perl -e 'use LWP::Simple;
print get "http://www.linuxjournal.com";'

Perl does not format the output for you. That's the difference between a web browser and a web client; the former is meant to retrieve information for humans, while the latter is meant to retrieve information for programs. In this particular case, we just want to simulate 100 retrievals of each of our programs via the web. Any timing differences will thus be due to the program on the server side—which, since they are identical, means that the differences will be due to mod_perl and Apache::DBI.

How much faster is the Apache::DBI version than its CGI counterpart? Here are the results I got running time-db.pl:

[1086] ~% ./time-db.pl
Benchmark: timing 100 iterations ...
Apache::DBI: 24 secs (1.77 usr 0.67 sys = 2.44 cpu)
Plain CGI: 394 secs (1.10 usr 0.61 sys = 1.71 cpu)

That's quite a difference. When I first ran this benchmark, I was convinced that the plain CGI program had somehow gotten stuck. Alas, that was not the case; the overhead associated with CGI was simply too great.

Here is a second run of the same benchmark, just for comparison.

[1099] ~% ./time-db.pl
Benchmark: timing 100 iterations ...
Apache::DBI: 28 secs (1.89 usr 0.61 sys = 2.50 cpu)
Plain CGI: 355 secs (1.15 usr 0.62 sys = 1.77 cpu)

Yes, it looks like CGI is indeed much slower. By the way, you can see that Apache::DBI used more CPU time than plain CGI—which means that the time was spent forking the new Perl process, rather than performing our program's computations.

What if we take out the Apache::DBI directive in srm.conf and restart the server? That would give us an indication of how much overhead was being used opening the database connection. As you can see, things do indeed slow down—although admittedly not by a huge amount:

[1104] ~% ./time-db.pl
Benchmark: timing 100 iterations ...
Apache::DBI: 34 secs (1.97 usr 0.63 sys = 2.60 cpu)
Plain CGI: 460 secs (1.19 usr 0.60 sys = 1.79 cpu)

The moral, then, seems to be that moving from CGI to mod_perl gives a huge performance boost, and that moving from DBI to Apache::DBI gives a moderate performance boost. The more database accesses your web applications do, the more useful these technologies will probably be in your work. Perl has always been known as a useful language, but rarely as one that can help you write fast software. Now, with mod_perl and Apache::DBI, you can write web applications quickly, and watch them run quickly as well.

Resources

Reuven M. Lerner is an Internet and Web consultant living in Haifa, Israel, who has been using the Web since early 1993. In his spare time, he cooks, reads and volunteers with educational projects in his community. You can reach him at reuven@netvision.net.il.

______________________

Webinar
One Click, Universal Protection: Implementing Centralized Security Policies on Linux Systems

As Linux continues to play an ever increasing role in corporate data centers and institutions, ensuring the integrity and protection of these systems must be a priority. With 60% of the world's websites and an increasing share of organization's mission-critical workloads running on Linux, failing to stop malware and other advanced threats on Linux can increasingly impact an organization's reputation and bottom line.

Learn More

Sponsored by Bit9

Webinar
Linux Backup and Recovery Webinar

Most companies incorporate backup procedures for critical data, which can be restored quickly if a loss occurs. However, fewer companies are prepared for catastrophic system failures, in which they lose all data, the entire operating system, applications, settings, patches and more, reducing their system(s) to “bare metal.” After all, before data can be restored to a system, there must be a system to restore it to.

In this one hour webinar, learn how to enhance your existing backup strategies for better disaster recovery preparedness using Storix System Backup Administrator (SBAdmin), a highly flexible bare-metal recovery solution for UNIX and Linux systems.

Learn More

Sponsored by Storix