httplog 2.1 review

httplog is a replacement for Apache's 'rotatelogs' and Andrew Ford's 'chronolog'

License: Free To Use But Restricted File size: 12K Developer: Eli Sand

0

httplog is a replacement for Apache's 'rotatelogs' and Andrew Ford's 'chronolog'. It allows you to specify a logfile using strftime paramaters in the filename to act as a template.

This means that the logs in your logfiles will also be sorted according to the filename. For example, if you specify a logfile of /var/log/http%Y%m%d.log, a new log file would be generated each day, with content for only that one day.

It also supports compression of logfiles using gzip, and many other useful functions.

It takes one argument at the command line, the path to a logfile as a template.
The template pathname can use any valid special characters permitted in strftime(3). Other optional arguments can be appended to the command line.

If you want to have httplog run as a different user and/or group (instead of running as root - possible security risks), you can use the `-u ' and `-g ' to have httplog run as that user and/or group.

You can also specify `-b ' to have httplog use a custom buffer size for memory storage of log data, rather than the default value of just line buffering mode (flushes data per newline).

If you want to have your logfiles gzipped because of size constraints, you can now use the `-z' option to have httplog gzip all logfiles once it's done writing all data to them (when a rollover period hits and a new logfile is to be created).

You may have a symlink to the currently active logfile by using the `-s' option and specifying a filename to be used as the symlink.

This program makes Apache act similar to Microsofts' IIS webserver in its logging style (as opposed to every logfile entry in one single huge logfile). This allows you to easily maintain your webserver logfiles for statistics packages in an easily organizable manner without user intervention.

Because of the buffer option added since v1.5, httplog may also be usefull for extremely large sites that get thousands of hits per second, and typically disable logging to save their hard drives from dying.

The buffer size option can be used to create an extremely large buffer in ram so that logfiles would not be flushed as often as if it were in line buffering mode, and in essence, could prolong the life of your valuable drives by decreasing their activity.

What's New in This Release:
Removed a couple of lines of code that could cause segfaults if errors were imminent while opening new a file or allocating the buffer memory