Sorry - that semicolon is an artifact of a change I made - I attempted
to put
postmaster_pid=$! inside the executing statement, but as I discovered,
if you put it before last &, you
get the backgrounded process before postmaster.
I'm presently checking out how the SLES init scripts work, but any
thoughts would be very welcome.
Thanks,
EE
Tom Lane wrote:
> Eric E <whalesuit@gmail.com> writes:
>
>> I chased the problem down to the following lines in pg_regress:
>>
>
>
>> "$bindir/postmaster" -D "$PGDATA" -F $postmaster_options
>> > "$LOGDIR/postmaster.log"; 2>&1 &
>>
> ^
>
> Is there really a semicolon there? There should not be (and isn't in
> my copy of 8.1)
>
> regards, tom lane
>
>