Hi Serge,

> However, there is a problem to your this patch. 
> The -q option of the monitor is lost.
> 
> -    runasowner -q "$OCF_RESKEY_psql $psql_options -c 'select now();'"
> +    runasowner "$OCF_RESKEY_psql $psql_options -c '$OCF_RESKEY_monitor_sql'"

The problem was only a -q option.

The additional patch of monitor_user,monitor_password,monitor_script which you 
made worked well. 

I contribute a patch corresponding to the change of the error loglevel of 
ocf_run for pgsql
successively from now on.

I contribute my patch by the weekend.

Best Regards,
Hideo Yamauchi.


--- [email protected] wrote:

> Hi Serge,
> 
> I confirm the movement of your patch from now on. 
> And I am going to contribute the patch of the change of loglevel of the 
> monitor processing at
> the time
> of the start.
> 
> However, there is a problem to your this patch. 
> The -q option of the monitor is lost.
> 
> -    runasowner -q "$OCF_RESKEY_psql $psql_options -c 'select now();'"
> +    runasowner "$OCF_RESKEY_psql $psql_options -c '$OCF_RESKEY_monitor_sql'"
> 
> I contribute a patch as soon as my patch making and confirmation work are 
> over.
> 
> Best Regards,
> Hideo Yamauchi.
> 
> 
> --- Serge Dubrouski <[email protected]> wrote:
> 
> > Attached is an updated version of the patch. monitor_script was
> > renamed to monitor_sql. Exit codes were updated as well.
> > 
> > On Tue, Oct 5, 2010 at 8:58 AM, Dejan Muhamedagic <[email protected]> 
> > wrote:
> > > On Fri, Oct 01, 2010 at 01:59:03PM -0600, Serge Dubrouski wrote:
> > >> Hello -
> > >>
> > >> Attached is a patch for pgsql RA that is supposed to improve monitor
> > >> operation by introducing 3 new variables: OCF_RESKEY_monitor_user,
> > >> OCF_RESKEY_monitor_password, OCF_RESKEY_monitor_script.
> > >
> > > Shouldn't the monitor_script be perhaps called monitor_code or
> > > something like that. Somehow, the script, to me at least, implies
> > > a file.
> > >
> > >> If they aren't
> > >> set RA fails back to the current behavior so it's backward compatible
> > >> and doesn't require changes for the current configuration if users
> > >> don't want to use these new features.
> > >
> > > Good.
> > >
> > >> Please review and push it into dev tip.
> > >
> > > The exit code on bad parameters should be changed to
> > > OCF_ERR_CONFIGURED. The difference to OCF_ERR_INSTALLED is that
> > > it means that the resource is going to fail on any node.
> > > OCF_ERR_INSTALLED means it can't run on this node.
> > >
> > > Cheers,
> > >
> > > Dejan
> > >
> > >> Thanks.
> > >>
> > >> --
> > >> Serge Dubrouski.
> > >
> > >
> > >> _______________________________________________________
> > >> Linux-HA-Dev: [email protected]
> > >> http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
> > >> Home Page: http://linux-ha.org/
> > >
> > > _______________________________________________________
> > > Linux-HA-Dev: [email protected]
> > > http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
> > > Home Page: http://linux-ha.org/
> > >
> > 
> > 
> > 
> > -- 
> > Serge Dubrouski.
> > > _______________________________________________________
> > Linux-HA-Dev: [email protected]
> > http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
> > Home Page: http://linux-ha.org/
> > 
> 
> _______________________________________________________
> Linux-HA-Dev: [email protected]
> http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
> Home Page: http://linux-ha.org/
> 

_______________________________________________________
Linux-HA-Dev: [email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/

Reply via email to