metabase: fix restoring backups

In practice, the restoration was always failing with errors like:

  pg_restore: [archiver (db)] could not execute query: ERROR: schema "???" already exists

because the database was first initialized, but the usual restore
scenario is:
 - instance is populated with initial data
 - data is replaced by restoring the data

this also change the timeout to 5 seconds, 30 is too much

software/metabase/buildout.hash.cfg

 [instance-profile]
 filename = instance.cfg.in
-md5sum = 0e74c862401f266111552b7a3611f7bf
+md5sum = c79f0bcf9dac25060340aa829c4f291c

software/metabase/instance.cfg.in

@@ -190,12 +190,13 @@ wrapper-path = $${directory:bin}/$${:_buildout_section_name_}
 command-line =
   sh -e -c "\
     echo 'This will replace current database with latest backup. Hit Ctrl+C to cancel';
-    sleep 30;
+    sleep 5;
     $${postgresql:bin}/pg_restore \
         --exit-on-error \
+        --clean \
+        --create \
         -h $${postgresql:pgdata-directory} \
         -U $${postgresql:superuser} \
-        -d $${postgresql:dbname} \
         $${postgresql-backup:backup-file}"
 
 [cron]