Apache Week
   
   Issue 161, 14th May 1999:  

Copyright ©2020 Red Hat, Inc

In this issue


Apache Status

Apache Site: www.apache.org
Release: 1.3.6 (Released 25th March 1999) (local download sites)
Beta: None

Apache 1.3.6 is the current stable release. Users of Apache 1.3.4 and earlier on Unix systems should upgrade to this version. Read the Guide to 1.3.6 for information about changes between 1.3.4 and 1.3.6 and between 1.2 and 1.3.6.

These bugs have been found in 1.3.6 and will be fixed in the next release.


Under Development

Patches for bugs in Apache 1.3.6 will be made available in the apply_to_1.3.6 subdirectory of the patches directory on the Apache site. Some new features and other unofficial patches are available in the 1.3 patches directory. For details of all previously reported bugs, see the Apache bug database and known bugs pages. Many common configuration questions are answered in the Apache FAQ.


Over three million Apache sites

The May Netcraft Server Survery shows Apache now powers over 3,000,000 sites on the Internet. Microsoft servers run 1,244,808 sites, and Nescape 356,207. Apache was also the only major server to increase its share of the server market, up by 1.03 to 57.22%. Microsoft was down 0.21 to 22.99% and Netscape down 0.20 to 6.58%. When servers based on Apache code are also included, Apache has to total share of 60.79%.


Apache refcards updated

The Apache and mod_perl reference cards at www.refcards.com have been updated. The Apache card now covers Apache 1.3.6, and the mod_perl card covers version 1.19. As usual, they are available as single-page or multi-page formats, in PDF or PS, and sized for printing onto A4 or US letter paper.


Apache in the news

PC Magazine tested four secure servers in Serve Your Site. The servers under test were Netscape Enterprise Server, Sun's Web Server, Microsoft IIS and Stronghold (an Apache derivative). The operating systems used where Windows NT, Solaris and Linux.

In the introduction, the fact that Apache is used on 55% of non-SSL sites is noted, but they add that this "overstates its popularity" because it is a count of unique domain names rather than unique servers. They note that in Netcraft's server of SSL sites, "Microsoft's 37 percent is almost twice that of Netscape's or Stronghold's share". They ignored the fact that secure Apache (that is, Apache plus mod_ssl or Apache-SSL) is now at 21%, and that if all Apache derived secure servers are counted they have about the same market share as IIS.

In the performance tests, the other servers easily beat Stronghold for serving static pages. There was no drop-off as the load increased (although interestingly there does appear to be a noticable drop-off in IIS performance at high loads). This is at odds with the results from MindCraft.

A more interesting test is the dynamic content one. Here they tested using CGI under Stronghold against ISAPI and NSAPI extensions to IIS and Netscape. Naturally the ISAPI and NSAPI versions performed much better than using CGI. It is strange that they did not test the various Apache methods of getting high-performance dynamic content, such as PHP, mod_perl or the Apache API (which is the direct equivalent of ISAPI and NSAPI). PHP even comes as standard with Stronghold, so they could easily have tested it in addition to the much slower CGI.

This bias is also apparent later in the article: "IIS and Netscape both offer alternatives to CGI in NSAPI and ISAPI, respectively--in-process API extensions that offer performance benefits. To achieve the same results with Apache, you must take your own native code programs and compile them as modules into the Apache Web server itself, a task that requires a lot more programming knowledge than writing ISAPI and NSAPI scripts does.". This is of course false: Apache offers many alternatives to CGI, ISAPI/NSAPI are not scripting languages and do require compilation and loading into the web server itself.

Apache/Stronghold actually did very well in the dynamic content test. If the unfair comparison with ISAPI and NSAPI extensions are ignored, Apache/Stronghold on Linux performed as well as IIS, and like the other servers showed no drop-off at high loads.