2009/2/25 Karsten Hilbert <Karsten(dot)Hilbert(at)gmx(dot)net>
>
> Think backwards: How would it automatically *dis*appear from
> there. Have the secondary create it and check for existance
> at regular intervals. Have the primary delete it at slightly
> shorter intervals.
>
> When the secondary finds it again after the check interval
> the primary didn't delete it. Wait some more for good
> measure and/or try to ping the primary some other way...
>
>
Thanks for the suggestion. I'm wondering what the chances of such a
create/delete system would have of accidentally causing failover? Is it
possible to use the archive command parameter to firstly run the pg_standby
command, and follow it by a touch /tmp/pg.triggerfile by using &&?