]> git.openstreetmap.org Git - nominatim.git/blobdiff - nominatim/indexer/indexer.py
Merge pull request #2475 from lonvia/catchup-mode
[nominatim.git] / nominatim / indexer / indexer.py
index 41535af8f2ae47bc7753d2cba1280fc617d4581a..50bd232e30dc0c341c6d43bbf8cf70eb720d37d0 100644 (file)
@@ -2,78 +2,82 @@
 Main work horse for indexing (computing addresses) the database.
 """
 import logging
-import select
+import time
 
 import psycopg2.extras
 
 from nominatim.indexer.progress import ProgressLogger
 from nominatim.indexer import runners
-from nominatim.db.async_connection import DBConnection
+from nominatim.db.async_connection import DBConnection, WorkerPool
 from nominatim.db.connection import connect
 
 LOG = logging.getLogger()
 
-class WorkerPool:
-    """ A pool of asynchronous database connections.
 
-        The pool may be used as a context manager.
+class PlaceFetcher:
+    """ Asynchronous connection that fetches place details for processing.
     """
-    REOPEN_CONNECTIONS_AFTER = 100000
+    def __init__(self, dsn, setup_conn):
+        self.wait_time = 0
+        self.current_ids = None
+        self.conn = DBConnection(dsn, cursor_factory=psycopg2.extras.DictCursor)
 
-    def __init__(self, dsn, pool_size):
-        self.threads = [DBConnection(dsn) for _ in range(pool_size)]
-        self.free_workers = self._yield_free_worker()
+        with setup_conn.cursor() as cur:
+            # need to fetch those manually because register_hstore cannot
+            # fetch them on an asynchronous connection below.
+            hstore_oid = cur.scalar("SELECT 'hstore'::regtype::oid")
+            hstore_array_oid = cur.scalar("SELECT 'hstore[]'::regtype::oid")
 
+        psycopg2.extras.register_hstore(self.conn.conn, oid=hstore_oid,
+                                        array_oid=hstore_array_oid)
 
-    def finish_all(self):
-        """ Wait for all connection to finish.
+    def close(self):
+        """ Close the underlying asynchronous connection.
         """
-        for thread in self.threads:
-            while not thread.is_done():
-                thread.wait()
+        if self.conn:
+            self.conn.close()
+            self.conn = None
 
-        self.free_workers = self._yield_free_worker()
 
-    def close(self):
-        """ Close all connections and clear the pool.
+    def fetch_next_batch(self, cur, runner):
+        """ Send a request for the next batch of places.
+            If details for the places are required, they will be fetched
+            asynchronously.
+
+            Returns true if there is still data available.
         """
-        for thread in self.threads:
-            thread.close()
-        self.threads = []
-        self.free_workers = None
+        ids = cur.fetchmany(100)
+
+        if not ids:
+            self.current_ids = None
+            return False
+
+        if hasattr(runner, 'get_place_details'):
+            runner.get_place_details(self.conn, ids)
+            self.current_ids = []
+        else:
+            self.current_ids = ids
 
+        return True
 
-    def next_free_worker(self):
-        """ Get the next free connection.
+    def get_batch(self):
+        """ Get the next batch of data, previously requested with
+            `fetch_next_batch`.
         """
-        return next(self.free_workers)
-
-
-    def _yield_free_worker(self):
-        ready = self.threads
-        command_stat = 0
-        while True:
-            for thread in ready:
-                if thread.is_done():
-                    command_stat += 1
-                    yield thread
-
-            if command_stat > self.REOPEN_CONNECTIONS_AFTER:
-                for thread in self.threads:
-                    while not thread.is_done():
-                        thread.wait()
-                    thread.connect()
-                ready = self.threads
-                command_stat = 0
-            else:
-                _, ready, _ = select.select([], self.threads, [])
+        if self.current_ids is not None and not self.current_ids:
+            tstart = time.time()
+            self.conn.wait()
+            self.wait_time += time.time() - tstart
+            self.current_ids = self.conn.cursor.fetchall()
 
+        return self.current_ids
 
     def __enter__(self):
         return self
 
 
     def __exit__(self, exc_type, exc_value, traceback):
+        self.conn.wait()
         self.close()
 
 
@@ -87,8 +91,19 @@ class Indexer:
         self.num_threads = num_threads
 
 
+    def has_pending(self):
+        """ Check if any data still needs indexing.
+            This function must only be used after the import has finished.
+            Otherwise it will be very expensive.
+        """
+        with connect(self.dsn) as conn:
+            with conn.cursor() as cur:
+                cur.execute("SELECT 'a' FROM placex WHERE indexed_status > 0 LIMIT 1")
+                return cur.rowcount > 0
+
+
     def index_full(self, analyse=True):
-        """ Index the complete database. This will first index boudnaries
+        """ Index the complete database. This will first index boundaries
             followed by all other objects. When `analyse` is True, then the
             database will be analysed at the appropriate places to
             ensure that database statistics are updated.
@@ -96,13 +111,10 @@ class Indexer:
         with connect(self.dsn) as conn:
             conn.autocommit = True
 
-            if analyse:
-                def _analyze():
+            def _analyze():
+                if analyse:
                     with conn.cursor() as cur:
                         cur.execute('ANALYZE')
-            else:
-                def _analyze():
-                    pass
 
             self.index_by_rank(0, 4)
             _analyze()
@@ -183,9 +195,6 @@ class Indexer:
                 total_tuples = cur.scalar(runner.sql_count_objects())
                 LOG.debug("Total number of rows: %i", total_tuples)
 
-                hstore_oid = cur.scalar("SELECT 'hstore'::regtype::oid")
-                hstore_array_oid = cur.scalar("SELECT 'hstore[]'::regtype::oid")
-
             conn.commit()
 
             progress = ProgressLogger(runner.name(), total_tuples)
@@ -194,49 +203,25 @@ class Indexer:
                 with conn.cursor(name='places') as cur:
                     cur.execute(runner.sql_get_objects())
 
-                    fetcher = DBConnection(self.dsn)
-                    psycopg2.extras.register_hstore(fetcher.conn,
-                                                    oid=hstore_oid,
-                                                    array_oid=hstore_array_oid)
-
-                    with WorkerPool(self.dsn, self.num_threads) as pool:
-                        places = self._fetch_next_batch(cur, fetcher, runner)
-                        while places is not None:
-                            if not places:
-                                fetcher.wait()
-                                places = fetcher.cursor.fetchall()
+                    with PlaceFetcher(self.dsn, conn) as fetcher:
+                        with WorkerPool(self.dsn, self.num_threads) as pool:
+                            has_more = fetcher.fetch_next_batch(cur, runner)
+                            while has_more:
+                                places = fetcher.get_batch()
 
-                            # asynchronously get the next batch
-                            next_places = self._fetch_next_batch(cur, fetcher, runner)
+                                # asynchronously get the next batch
+                                has_more = fetcher.fetch_next_batch(cur, runner)
 
-                            # And insert the curent batch
-                            for idx in range(0, len(places), batch):
-                                worker = pool.next_free_worker()
-                                part = places[idx:idx+batch]
-                                LOG.debug("Processing places: %s", str(part))
-                                runner.index_places(worker, part)
-                                progress.add(len(part))
+                                # And insert the curent batch
+                                for idx in range(0, len(places), batch):
+                                    part = places[idx:idx + batch]
+                                    LOG.debug("Processing places: %s", str(part))
+                                    runner.index_places(pool.next_free_worker(), part)
+                                    progress.add(len(part))
 
-                            places = next_places
-
-                        pool.finish_all()
-
-                    fetcher.wait()
-                    fetcher.close()
+                            LOG.info("Wait time: fetcher: %.2fs,  pool: %.2fs",
+                                     fetcher.wait_time, pool.wait_time)
 
                 conn.commit()
 
         progress.done()
-
-
-    def _fetch_next_batch(self, cur, fetcher, runner):
-        ids = cur.fetchmany(1000)
-
-        if not ids:
-            return None
-
-        if not hasattr(runner, 'get_place_details'):
-            return ids
-
-        runner.get_place_details(fetcher, ids)
-        return []