Speed up SQLite unit test CI

changelog.d/15334.misc

@@ -0,0 +1 @@
+Speed up unit tests when using SQLite3.

synapse/storage/engines/sqlite.py

@@ -34,6 +34,13 @@ def __init__(self, database_config: Mapping[str, Any]):
             ":memory:",
         )
 
+        # A connection to a database that has already been prepared, to use as a
+        # base for an in-memory connection. This is used during unit tests to
+        # speed up setting up the DB.
+        self._prepped_conn: Optional[sqlite3.Connection] = database_config.get(
+            "_TEST_PREPPED_CONN"
+        )
+
         if platform.python_implementation() == "PyPy":
             # pypy's sqlite3 module doesn't handle bytearrays, convert them
             # back to bytes.
@@ -84,7 +91,15 @@ def on_new_connection(self, db_conn: "LoggingDatabaseConnection") -> None:
             # In memory databases need to be rebuilt each time. Ideally we'd
             # reuse the same connection as we do when starting up, but that
             # would involve using adbapi before we have started the reactor.
-            prepare_database(db_conn, self, config=None)
+            #
+            # If we have a `prepped_conn` we can use that to initialise the DB,
+            # otherwise we need to call `prepare_database`.
+            if self._prepped_conn is not None:
+                # Initialise the new DB from the pre-prepared DB.
+                assert isinstance(db_conn.conn, sqlite3.Connection)
+                self._prepped_conn.backup(db_conn.conn)
+            else:
+                prepare_database(db_conn, self, config=None)
 
         db_conn.create_function("rank", 1, _rank)
         db_conn.execute("PRAGMA foreign_keys = ON;")

tests/server.py

@@ -16,6 +16,7 @@
 import logging
 import os
 import os.path
+import sqlite3
 import time
 import uuid
 import warnings
@@ -79,7 +80,9 @@
 from synapse.logging.context import ContextResourceUsage
 from synapse.server import HomeServer
 from synapse.storage import DataStore
+from synapse.storage.database import LoggingDatabaseConnection
 from synapse.storage.engines import PostgresEngine, create_engine
+from synapse.storage.prepare_database import prepare_database
 from synapse.types import ISynapseReactor, JsonDict
 from synapse.util import Clock
 
@@ -104,6 +107,10 @@
 # the type of thing that can be passed into `make_request` in the headers list
 CustomHeaderType = Tuple[Union[str, bytes], Union[str, bytes]]
 
+# A pre-prepared SQLite DB that is used as a template when creating new SQLite
+# DB each test run. This dramatically speeds up test set up when using SQLite.
+PREPPED_SQLITE_DB_CONN: Optional[LoggingDatabaseConnection] = None
+
 
 class TimedOutException(Exception):
     """
@@ -899,6 +906,22 @@ def setup_test_homeserver(
             "args": {"database": test_db_location, "cp_min": 1, "cp_max": 1},
         }
 
+        # Check if we have set up a DB that we can use as a template.
+        global PREPPED_SQLITE_DB_CONN
+        if PREPPED_SQLITE_DB_CONN is None:
+            temp_engine = create_engine(database_config)
+            PREPPED_SQLITE_DB_CONN = LoggingDatabaseConnection(
+                sqlite3.connect(":memory:"), temp_engine, "PREPPED_CONN"
+            )
+
+            database = DatabaseConnectionConfig("master", database_config)
+            config.database.databases = [database]
+            prepare_database(
+                PREPPED_SQLITE_DB_CONN, create_engine(database_config), config
+            )
+
+        database_config["_TEST_PREPPED_CONN"] = PREPPED_SQLITE_DB_CONN
+
     if "db_txn_limit" in kwargs:
         database_config["txn_limit"] = kwargs["db_txn_limit"]
 

tests/unittest.py

@@ -146,6 +146,9 @@ def setUp(orig: Callable[[], R]) -> R:
                     % (current_context(),)
                 )
 
+            # Disable GC for duration of test. See below for why.
+            gc.disable()
+
             old_level = logging.getLogger().level
             if level is not None and old_level != level:
 
@@ -163,12 +166,19 @@ def tearDown(orig: Callable[[], R]) -> R:
 
             return orig()
 
+        # We want to force a GC to workaround problems with deferreds leaking
+        # logcontexts when they are GCed (see the logcontext docs).
+        #
+        # The easiest way to do this would be to do a full GC after each test
+        # run, but that is very expensive. Instead, we disable GC (above) for
+        # the duration of the test so that we only need to run a gen-0 GC, which
+        # is a lot quicker.
+
         @around(self)
         def tearDown(orig: Callable[[], R]) -> R:
             ret = orig()
-            # force a GC to workaround problems with deferreds leaking logcontexts when
-            # they are GCed (see the logcontext docs)
-            gc.collect()
+            gc.collect(0)
+            gc.enable()
             set_current_context(SENTINEL_CONTEXT)
 
             return ret