in reply to Re: Tracking actual time spent sleeping: Is this bad coding?
in thread Tracking actual time spent sleeping: Is this bad coding?

Dunno about the OP but I have a daemonized process that monitors some hardware and sleeps a lot. I do not use cron though because the start up costs are high compared to the actual work it does.

Cheers,
R.

Pereant, qui ante nos nostra dixerunt!
  • Comment on Re^2: Tracking actual time spent sleeping: Is this bad coding?

Replies are listed 'Best First'.
Re^3: Tracking actual time spent sleeping: Is this bad coding?
by friedo (Prior) on Feb 02, 2005 at 15:25 UTC
    If you're worried about the startup costs of cron, I've had excellent experiences with Schedule::Cron. You can daemonize the process and tell it to run a coderef on schedule.
[reply]
Re^3: Tracking actual time spent sleeping: Is this bad coding?
by jpk236 (Monk) on Feb 02, 2005 at 15:36 UTC
    I'm unfamiliar with any start-up costs of cron. I use it quite frequently on most boxes I administer. Would you mind going into detail? Perhaps after your explanation I'll change my ways.

    - Justin
[reply]

      Hi jpk236,

      It is not a cost of cron but the costs of starting my process that is high. My Process has to load and compile about 700 regular expressions each time it starts. Why do this every N minutes when you can do it once and hold the results in memory ?

      I do like the look of Schedule::Cron and it looks like it would provide a nice solution to the OP's problem, friedo++

      Cheers,
      R.

      Pereant, qui ante nos nostra dixerunt!
[reply]

      There are no start-up costs with cron itself. The start-up costs referred to in this thread are associated with the Perl code and would be the same whether the program is ran from cron or from the command line (e.g compiling modules, connecting to databases etc).

      -- vek --
[reply]

In Section Seekers of Perl Wisdom