hi, thanks for your prompt answer (i am only parttime worker for this job, therefore sorry for my late reaction).
if insserv is the way to go, than I will use it also for checks. I found a misconfiguration with check-initd-order, where insserv didn't complain. I cannot remember the situation, so I am not able to describe it, but next time I will take care of it. btw, if I am disabling a script(temporary) with e.g. (debian squeeze): insserv -v -f apache2,stop=2 than start and kill scripts are existing lrwxrwxrwx 1 root root 17 3. Mar 04:57 /etc/rc2.d/K01apache2 -> ../init.d/apache2 lrwxrwxrwx 1 root root 17 25. Feb 06:43 /etc/rc2.d/S19apache2 -> ../init.d/apache2 is this the indented behaviour, or a bug? (yes, the man page says "List of scripts which have to be added to the specified runlevels to be started with" and for stop I will "remove"). but it is misleading and also in being conservative and not willing to break systems (which i really prefer!), it doesn't make sense not to remove the start scripts. so I am teaching my colleagues to always use insserv -v -r apache2 insserv -v -f apache2,stop=0,1,2,3,4,5,6 to disable, and insserv -v -r apache2 insserv -v -d apache2 to enable it, as a general rule for deactivating/activating skripts. and to be realy safe I recommand to use insserv -v -r apache2 insserv -v -r apache2 insserv -v -f apache2,stop=0,1,2,3,4,5,6 because if (by a mistake of someone) a start and a kill script exists in one runlevel, than the first "insserv -v -r apache2" will only remove the kill script and the start script will still exist after the stop command and start the service (e.g. after a reboot some time in the future). please let me know if you are not interested in this sort of feedback and excuse me for bother you. andreas > -----Original Message----- > From: Petter Reinholdtsen [mailto:[email protected]] > Sent: Donnerstag, 12. Juli 2012 19:43 > To: Unterluggauer Andreas, ON-T > Subject: Re: check-initd-order uses hardcoded runlevel 2 for > boot check > > [Andreas Unterluggauer] > > hi pere, > > Hi. > > > the useful script /usr/share/insserv/check-initd-order uses the > > hardcoded runlevel 2 for checking the boot sequence. This is > > misleading if someone (like me) uses an other boot runlevel in the > > inittab. > > Yes, that is correct. :) > > The script does some checking, but not all the checking that could be > done. It did a good job when we needed to track down and fix a lot of > dependency problems in Debian, but these days the init.d script > dependnecies are so good that it probably is better to use insserv > directly to test dependencies. But please go ahead if you want to > send patches to improve the script. > > -- > Happy hacking > Petter Reinholdtsen > _______________________________________________ Pkg-sysvinit-devel mailing list [email protected] http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-sysvinit-devel
