Forum OpenACS Q&A: Slackware 9.0 init scripts

Collapse
Posted by Brad Ford on
I'm following the installation guide for 4.6.x and have finished compiling postgres. I had to go with slackware due to problems with my motherboard (ecs all in one duron) - redhat and mandrake had some 'issues'. I'm at the point of setting up the init scripts. The docs have references for red hat, debian and suse. Slackware is SysV compatible (I think) so I think the red hat instructions will work. Anyone else out there with Slackware experience that can point me in the right direction for init scripts?
Collapse
Posted by Brad Ford on
Answering my own question: I did some playing - turns out slackware uses a bsd style rc.M script normally but also has an rc.sysvinit script for sysV compatability. I didn't want to get into playing with the rc.M file (looks like just cutting and pasting the proper start/stop sections from the postgresql.txt file into the appropriate place in the rc.M file) so I read through the rc.sysvinit file to figure out where to put the K20 and S20 symlinks to /etc/rc.d/init.d/postgresql - first you have to create the /etc/rc.d/init.d directory, then copy postgresql into it. Next, create the symlinks as described in the Suse section of the documentation (ln -s /etc/rc.d/init.d/postgresql K20postgresql, then S20postgresql). Next, create rc2.d, rc3.d, rc4.d, rc5.d directories in /etc/rc.d. Finally, copy the K20 and S20postgresql symlinks into each of those new directories. et voila - it worked. Probably not the most elegant way for Slackware but until I learn a bit more...