David A. Bandel wrote: > What candidates should _really_ know is how to figure out > the bootup sequence, be it via SysV style scripts, startup, > rc scripts (a la Slackware), etc; i.e., how does the system > boot? > Problem becomes, how to test for _that_ kind of > knowledge. So we expect them to know S## and K## > startup sequences and how to manipulate them (via > chkconfig, update-* scripts, etc.).
When in doubt, LSB. Of course, LSB doesn't address everything. Then you have these issues: For one, start-up sequence is no long sequential in some systems. More and more are going parallel. Although LSB does attempt to account for such with its headers for SysV startup, it doesn't address implementation aspects, advanced dependency resolution between services, etc... Secondly, the administrative commands for managing init may vary by solution. LSB doesn't even address this, even though many POSIX systems standardized on "service" over a decade ago - making it agnostic of the actual SysV init system. Sometimes some of us suggest things only to be berrated as "oh, that's a Red Hat'ism," even if it's used by more than just Fedora, CentOS and RHEL, sometimes even older, previous Red Hat released conform, Ubuntu grabs a lot from Fedora (and vice-versa, like upstart), and then out to non- Linux like Solaris, etc... Given Red Hat's marketshare for replacing Sun, it's very commona, but hardly does Red Hat do it alone. And I've said "I told you so" too many times to even bother when people use the phrase. -- Bryan J Smith - mailto:[email protected] http://www.linkedin.com/in/bjsmith Sent via BlackBerry from T-Mobile _______________________________________________ lpi-examdev mailing list [email protected] http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-examdev
