On Thu, Feb 12, 2009 at 04:12:07PM +0100, Paul Bijnens wrote:
On 2009-02-12 15:47, jkinz@kinz.org wrote:
On Thu, Feb 12, 2009 at 01:08:24PM +0800, Fajar Priyanto wrote:
Hi Fajar, re logrotate and crontab: I was speaking from instinct knowing that crontab and a simple script to do the actual rotation is all that is needed. What I didn't expect was that someone had actually reinvented all the functionality of crontab to create a totally new utility that is much more limited than crontab.
You mean "anacron"? Then you must read the man page again.
No, I mean logrotate.
logrotate still is that simple script to be invoked by (ana)cron.
No, logrotate is not a script.(It should be.)
Instead it is a complete stand alone utility written in C. In the version I just built from source, the executable is 65K bytes in size. It recreates most of what cron does internally to see if it needs to actually do anything during its once daily invocation. It is well written but I think the decision to create it was a flawed one, re-inventing the wheel where a script would have been OK. Even a script that allowed the same functionality as logrotate except for the parts done by cron would be fine.
On Centos/RHEL: (4.4) # file $(which logrotate) /usr/sbin/logrotate: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), for GNU/Linux 2.2.5, dynamically linked (uses shared libs), stripped
On that "other" user friendly distro.... :-) (LTS 6.06) # file $(which logrotate) /usr/sbin/logrotate: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), for GNU/Linux 2.2.0, dynamically linked (uses shared libs), for GNU/Linux 2.2.0, stripped