From: Andreas E. <ae...@op...> - 2004-10-26 08:51:34
|
Stanley Hopcroft wrote: > Dear Sir, > > I am writing to thank you for your letter and say, > > On Wed, Oct 20, 2004 at 09:01:28PM -0500, Ethan Galstad wrote: > >>Following up on an old post that needs addressing... >> >>The info from the Perl docs you provided earlier state that lost >>memory can only be freed by stopping the process. > > > Agreed. > > >>Rather than >>including this functionality in Nagios (which would be a bit kludgy), >>I would recommend another approach (which is still hackish, but >>better IMO)... >> > > > > >>Use cron to periodically run a script that calls the check_vsz plugin >>to determine the memory usage of the Nagios daemon. If the plugin >>returns a critical state, restart Nagios using the init script. >>Simple, but effective. >> >> > > > Yes, this is a much better way (even periodic restarts from cron without > checking mem use would be sufficient). > Umm.. How about the old-fashioned and cleaner way, using vfork(2) and execve(2)? -- Andreas Ericsson and...@op... OP5 AB www.op5.se Lead Developer |