Revert to using primary database in IndexingScheduler (#26754)

app/workers/scheduler/indexing_scheduler.rb

@@ -16,9 +16,7 @@ class Scheduler::IndexingScheduler
     indexes.each do |type|
       with_redis do |redis|
         redis.sscan_each("chewy:queue:#{type.name}", count: SCAN_BATCH_SIZE).each_slice(IMPORT_BATCH_SIZE) do |ids|
-          with_read_replica do
+          type.import!(ids)
-            type.import!(ids)
-          end
 
           redis.srem("chewy:queue:#{type.name}", ids)
         end