I know, that just sounds wrong. However, I have to get around the problem. We're setting up an apache-installation in a HP ServiceGuard cluster and ran into a problem. There's virtual hosts tied to the ip-numbers jumping around in the cluster, so each virtual host gets it's own config file, like /cluster_pkg1/conf/httpd.conf /cluster_pkg2/conf/httpd.conf These files are included in the main httpd.conf like this Include /cluster_conf/httpd-confs/ In this directory there's a bunch of symlinks pointing to the httpd.conf-files in the cluster packages, like this /cluster_conf/httpd-confs/clpkg1.conf -> /cluster_pkg1/conf/httpd.conf /cluster_conf/httpd-confs/clpkg2.conf -> /cluster_pkg2/conf/httpd.conf Now, when one of these symlinks are broken apache will not start. The symlinks are broken when the cluster package is running on some other node since /clpkg1 and /clpkg2 are just empty mountpoints when the package has moved. Is there a way around this? There aren't any other ways I can think of to do this other than some shellscript to generate a new httpd.conf and restart apache at each clusterpackage event, which would be sort of clumsy. We are using 1.3.27 at the moment. -- There are two major products that come out of Berkeley: LSD and UNIX. We don't believe this to be a coincidence. -- Jeremy S. Anderson Thomas Johansson | http://thomas.watchmen.se --------------------------------------------------------------------- The official User-To-User support forum of the Apache HTTP Server Project. See <URL:http://httpd.apache.org/userslist.html> for more info. To unsubscribe, e-mail: users-unsubscribe@xxxxxxxxxxxxxxxx " from the digest: users-digest-unsubscribe@xxxxxxxxxxxxxxxx For additional commands, e-mail: users-help@xxxxxxxxxxxxxxxx
|