Hot standby won't start

On Mon, Oct 10, 2011 at 11:30 AM, Brandon Phelps <bphelps@gls.com> wrote:

Hello all,

I have written a script which stops postgres on my standby server, executes
pg_start_backup on the master, rsync's the data directory and tablespaces
over to the standby, executes pg_stop_backup on the master, and attempts to
then start postgresql on the standby.  My problem is the following error
when I try to start postgresql on the standby:

 * Starting PostgreSQL 9.1 database server
 * The PostgreSQL server failed to start. Please check the log output:
2011-10-10 12:25:21 EDT LOG:  database system was shut down at 2011-10-10
12:25:18 EDT
2011-10-10 12:25:21 EDT LOG:  entering standby mode
2011-10-10 12:25:21 EDT LOG:  streaming replication successfully connected
to primary
2011-10-10 12:25:21 EDT LOG:  incomplete startup packet
2011-10-10 12:25:21 EDT WARNING:  WAL was generated with wal_level=minimal,
data may be missing
2011-10-10 12:25:21 EDT HINT:  This happens if you temporarily set
wal_level=minimal without taking a new base backup.
2011-10-10 12:25:21 EDT FATAL:  hot standby is not possible because
wal_level was not set to "hot_standby" on the master server
2011-10-10 12:25:21 EDT HINT:  Either set wal_level to "hot_standby" on the
master, or turn off hot_standby here.
2011-10-10 12:25:21 EDT LOG:  startup process (PID 20370) exited with exit
code 1
2011-10-10 12:25:21 EDT LOG:  aborting startup due to startup process
failure

I have tried running pg_resetxlog on both servers, and have verified that
pg_hba.conf contains my replication user.  I have also verified that the
password I am supplying in my recovery.conf is correct.

Any ideas why I would get this message?  I know that wal_level = hot_standby
on the master and the slave.

did you set that *before* setting up the standby or after? did you
remember to u-comment the line in postgresql.conf?

merlin