Re: cron problem with Oracle's lsnrctl and MySQL

From: Daryl (dkjjr_at_ohio.net)
Date: 02/03/04 

Date: 3 Feb 2004 12:49:18 -0800

Pnoi <pnoi@kremlin.nyet.ru> wrote in message news:<xWyTb.31451$dC7.26908@newssvr33.news.prodigy.com>...
> On Mon, 02 Feb 2004 05:05:55 -0800, Daryl enlightened everyone by
> writting:
>
> > Pnoi <pnoi@kremlin.nyet.ru> wrote in message news:<z3nTb.2782$iW2.2400@newssvr31.news.prodigy.com>...
> >> On Sat, 31 Jan 2004 18:28:41 -0800, Daryl enlightened everyone by
> >> writting:
> >>
> >> > All,
> >> >
> >> > I am trying to launch a backup from cron that, among other things,
> >> > backs up Oracle and MySQL data. A simplified script looks like:
> >> >
> >> > #! /bin/sh
> >> >
> >> > su - oracle8i -c "lsnrctl stop"
> >> > # Do some exporting and gzipping, etc.
> >> > su - oracle8i -c "lsnrctl start"
> >> >
> >> > the last line causes the script to become defunct such that the script
> >> > output is not mailed and my process list looks something like:
> >> >
> >> > root 6554 31375 CROND
> >> > root 6555 6554 [sh <defunct>]
> >> > root 6559 6554 [sendmail]
> >> >
> >> > If I comment out the starting of the listener the script completes
> >> > just fine. I also have the same issue trying to start MySQL (same
> >> > fashion - stop MySQL, backup data, start MySQL). The full script runs
> >> > just fine outside of cron. I am running:
> >> >
> >> > Red Hat Linux 9 on an AMD box with the included Vixie cron
> >> > Oracle 8i
> >> > MySQL 4.0.15
> >> >
> >> > Has anyone seen this behavior before? Does anyone know if I am doing
> >> > something wrong? For now I can manually start these items in the
> >> > morning after the back ups complete so I can view (be mailed) the log
> >> > but I would like to resolve this issue.
> >> >
> >>
> >> INAE would exec 'su - oracle8i -c "lsnrctl start"' or
> >> source file_containing_start_command work ?
> >>
> >> > Thank you kindly,
> >> > Daryl
> >
> >
> > Thank you, but sourcing does not make a difference. BTW, I am not
> > familiar with INAE - what is it?
> >
> > Thanks!
>
> I asked the same question -IANE = I am no expert
>
> I guess placing the start commands into a separate script that the
> backup script calls is the same as source.
>
> How about $(su - oracle8i -c "lsnrctl start") ?

That did not work either. I contacted Paul Vixie who responded quite
promptly and said to try two things:

1. A new version at ftp.isc.org/isc/cron - which did not work
2. su - oracle8i -c "lsnrctl start >/dev/null 2>&1 </dev/null &" -
which DID work.

Thank you Pnoi for your time trying to work through this with me.
Thanks also to Paul for his time.

Daryl

Relevant Pages

  • Re: cron problem with Oracles lsnrctl and MySQL
    ... but sourcing does not make a difference. ... BTW, I am not ... > familiar with INAE - what is it? ... I guess placing the start commands into a separate script that the ...
    (comp.os.linux)
  • Re: cron problem with Oracles lsnrctl and MySQL
    ... >> I guess placing the start commands into a separate script that the ... >> backup script calls is the same as source. ... I contacted Paul Vixie who responded quite ...
    (comp.os.linux)