Merge pull request #612 from matrix-org/erikj/cache_size

Add environment variable SYNAPSE_CACHE_FACTOR, default it to 0.1

README.rst

@@ -566,3 +566,20 @@ Building internal API documentation::
 
     python setup.py build_sphinx
 
+
+
+Halp!! Synapse eats all my RAM!
+===============================
+
+Synapse's architecture is quite RAM hungry currently - we deliberately
+cache a lot of recent room data and metadata in RAM in order to speed up
+common requests.  We'll improve this in future, but for now the easiest
+way to either reduce the RAM usage (at the risk of slowing things down)
+is to set the almost-undocumented ``SYNAPSE_CACHE_FACTOR`` environment
+variable.  Roughly speaking, a SYNAPSE_CACHE_FACTOR of 1.0 will max out
+at around 3-4GB of resident memory - this is what we currently run the
+matrix.org on.  The default setting is currently 0.1, which is probably
+around a ~700MB footprint.  You can dial it down further to 0.02 if
+desired, which targets roughly ~512MB.  Conversely you can dial it up if
+you need performance for lots of users and have a box with a lot of RAM.
+

jenkins.sh

@@ -3,6 +3,7 @@
 : ${WORKSPACE:="$(pwd)"}
 
 export PYTHONDONTWRITEBYTECODE=yep
+export SYNAPSE_CACHE_FACTOR=1
 
 # Output test results as junit xml
 export TRIAL_FLAGS="--reporter=subunit"

synapse/handlers/presence.py

@@ -130,6 +130,10 @@ class PresenceHandler(BaseHandler):
             for state in active_presence
         }
 
+        metrics.register_callback(
+            "user_to_current_state_size", lambda: len(self.user_to_current_state)
+        )
+
         now = self.clock.time_msec()
         for state in active_presence:
             self.wheel_timer.insert(

synapse/handlers/receipts.py

@@ -36,8 +36,6 @@ class ReceiptsHandler(BaseHandler):
         )
         self.clock = self.hs.get_clock()
 
-        self._receipt_cache = None
-
     @defer.inlineCallbacks
     def received_client_receipt(self, room_id, receipt_type, user_id,
                                 event_id):

synapse/util/caches/descriptors.py

@@ -28,6 +28,7 @@ from twisted.internet import defer
 
 from collections import OrderedDict
 
+import os
 import functools
 import inspect
 import threading
@@ -38,6 +39,9 @@ logger = logging.getLogger(__name__)
 _CacheSentinel = object()
 
 
+CACHE_SIZE_FACTOR = float(os.environ.get("SYNAPSE_CACHE_FACTOR", 0.1))
+
+
 class Cache(object):
 
     def __init__(self, name, max_entries=1000, keylen=1, lru=True, tree=False):
@@ -140,6 +144,8 @@ class CacheDescriptor(object):
     """
     def __init__(self, orig, max_entries=1000, num_args=1, lru=True, tree=False,
                  inlineCallbacks=False):
+        max_entries = int(max_entries * CACHE_SIZE_FACTOR)
+
         self.orig = orig
 
         if inlineCallbacks:

synapse/util/caches/stream_change_cache.py

@@ -18,11 +18,15 @@ from synapse.util.caches import cache_counter, caches_by_name
 
 from blist import sorteddict
 import logging
+import os
 
 
 logger = logging.getLogger(__name__)
 
 
+CACHE_SIZE_FACTOR = float(os.environ.get("SYNAPSE_CACHE_FACTOR", 0.1))
+
+
 class StreamChangeCache(object):
     """Keeps track of the stream positions of the latest change in a set of entities.
 
@@ -33,7 +37,7 @@ class StreamChangeCache(object):
     old then the cache will simply return all given entities.
     """
     def __init__(self, name, current_stream_pos, max_size=10000, prefilled_cache={}):
-        self._max_size = max_size
+        self._max_size = int(max_size * CACHE_SIZE_FACTOR)
         self._entity_to_key = {}
         self._cache = sorteddict()
         self._earliest_known_stream_pos = current_stream_pos

tests/storage/test_appservice.py

@@ -35,7 +35,8 @@ class ApplicationServiceStoreTestCase(unittest.TestCase):
     def setUp(self):
         self.as_yaml_files = []
         config = Mock(
-            app_service_config_files=self.as_yaml_files
+            app_service_config_files=self.as_yaml_files,
+            event_cache_size=1,
         )
         hs = yield setup_test_homeserver(config=config)
 
@@ -109,7 +110,8 @@ class ApplicationServiceTransactionStoreTestCase(unittest.TestCase):
         self.as_yaml_files = []
 
         config = Mock(
-            app_service_config_files=self.as_yaml_files
+            app_service_config_files=self.as_yaml_files,
+            event_cache_size=1,
         )
         hs = yield setup_test_homeserver(config=config)
         self.db_pool = hs.get_db_pool()
@@ -438,7 +440,7 @@ class ApplicationServiceStoreConfigTestCase(unittest.TestCase):
         f1 = self._write_config(suffix="1")
         f2 = self._write_config(suffix="2")
 
-        config = Mock(app_service_config_files=[f1, f2])
+        config = Mock(app_service_config_files=[f1, f2], event_cache_size=1)
         hs = yield setup_test_homeserver(config=config, datastore=Mock())
 
         ApplicationServiceStore(hs)
@@ -448,7 +450,7 @@ class ApplicationServiceStoreConfigTestCase(unittest.TestCase):
         f1 = self._write_config(id="id", suffix="1")
         f2 = self._write_config(id="id", suffix="2")
 
-        config = Mock(app_service_config_files=[f1, f2])
+        config = Mock(app_service_config_files=[f1, f2], event_cache_size=1)
         hs = yield setup_test_homeserver(config=config, datastore=Mock())
 
         with self.assertRaises(ConfigError) as cm:
@@ -464,7 +466,7 @@ class ApplicationServiceStoreConfigTestCase(unittest.TestCase):
         f1 = self._write_config(as_token="as_token", suffix="1")
         f2 = self._write_config(as_token="as_token", suffix="2")
 
-        config = Mock(app_service_config_files=[f1, f2])
+        config = Mock(app_service_config_files=[f1, f2], event_cache_size=1)
         hs = yield setup_test_homeserver(config=config, datastore=Mock())
 
         with self.assertRaises(ConfigError) as cm: