Bernd Schemmer wrote: After takeing a look at the current implementation of legacy scripts (AFAIK they are executed by the scripts /sbin/rc*) I think it should be not to difficult to implement a restart of the milestones using a (temporary) smf attribute for the last executed legacy script.
Should I create an RFE for this ? Or do I miss something? regards Bernd >If a legacy script fails (e.g. the timeout value of the milestone service is >reached) the SMF daemon kills the script and sets the status of the milestone >service to maintenance. The legacy scripts in that runlevel that would run >after the failed script are not executed. > >I'm not sure if this is a useful behaviour. It's not compatible with the >previous Solaris version and I don't know why it's implemented this way. > >Another question is: How to clear this situation? > >One point is to find & fix the failed legacy script, but how to set the >service status to "online" ? > >"svcadm clear FRMI" works but doing this will re-execute *all* legacy scripts >in the runlevel and that's normally not what I want to do . > >One method I can think about, is to temporary rename the already executed >legacy scripts before doing the svcadm refresh but that's not really a >solution. > > >regards > >Bernd > > >This message posted from opensolaris.org >_______________________________________________ >smf-discuss mailing list >smf-discuss at opensolaris.org > > > > -- Bernd Schemmer, Frankfurt am Main, Germany http://home.arcor.de/bnsmb/index.html M?s temprano que tarde el mundio cambiar? Fidel Castro
