Posted by voronwae on 04/03/06 22:56

Hi folks. Either I'm missing something really obvious (most likely) or
I'm missing something really subtle. I've been building up a machine
as an IMP webmail server, with php 5.1.2, cyrus-sasl, openldap, imap,
fileinfo, mod_perl, sqlite, horde, Berkeley DB and PostgreSQL, among
others, on top of an old Potatoe machine that was already alive and in
production so I just went with it. Everything worked great up until
the installation of fileinfo, which I abruptly decided to install after
everything else was up and running and Horde's test.php screen was
already up and displaying features.

At that point (after installing fileinfo with PECL) Apache 1.3.34
bailed on me, refusing to start (or restart) with no error message
whatsever. I can't find a trace of an error message in either Apache's
logs or any of the system logs. If I comment out PHP's LoadModule and
AddModule lines in my httpd.conf, Apache starts just fine. If I
uncomment them, Apache refuses to start. "Apachectl start" comes back
and says that httpd is up and running, but "ps -aux" doesn't show any
Apache processes. I had initially started with the latest stable
snapshot of PHP, but I blew away /usr/local/lib/php and reinstalled
5.1.2 just in case that was the problem, with identical results.

Somewhere on this system is file that's poisonous, I think, but what
gets me is that I don't even get any error logs. Does anybody have any
ideas for troubleshooting?

Thanks in advance!
voronwae

[Back to original message]