from __future__ import absolute_import from functools import wraps from django.core.cache import cache as djcache from django.core.cache import get_cache from django.conf import settings from zerver.lib.utils import statsd, statsd_key, make_safe_digest import time import base64 import random import sys import os import os.path import hashlib memcached_time_start = 0 memcached_total_time = 0 memcached_total_requests = 0 def get_memcached_time(): return memcached_total_time def get_memcached_requests(): return memcached_total_requests def memcached_stats_start(): global memcached_time_start memcached_time_start = time.time() def memcached_stats_finish(): global memcached_total_time global memcached_total_requests global memcached_time_start memcached_total_requests += 1 memcached_total_time += (time.time() - memcached_time_start) def get_or_create_key_prefix(): if settings.TEST_SUITE: # This sets the prefix mostly for the benefit of the JS tests. # The Python tests overwrite KEY_PREFIX on each test. return 'test_suite:' + str(os.getpid()) + ':' filename = os.path.join(settings.DEPLOY_ROOT, "memcached_prefix") try: fd = os.open(filename, os.O_CREAT | os.O_EXCL | os.O_RDWR, 0444) prefix = base64.b16encode(hashlib.sha256(str(random.getrandbits(256))).digest())[:32].lower() + ':' # This does close the underlying file with os.fdopen(fd, 'w') as f: f.write(prefix + "\n") except OSError: # The file already exists tries = 1 while tries < 10: with file(filename, 'r') as f: prefix = f.readline()[:-1] if len(prefix) == 33: break tries += 1 prefix = '' time.sleep(0.5) if not prefix: sys.exit("Could not read memcache key prefix file") return prefix KEY_PREFIX = get_or_create_key_prefix() def bounce_key_prefix_for_testing(test_name): global KEY_PREFIX KEY_PREFIX = test_name + ':' + str(os.getpid()) + ':' def get_cache_backend(cache_name): if cache_name is None: return djcache return get_cache(cache_name) def cache_with_key(keyfunc, cache_name=None, timeout=None, with_statsd_key=None): """Decorator which applies Django caching to a function. Decorator argument is a function which computes a cache key from the original function's arguments. You are responsible for avoiding collisions with other uses of this decorator or other uses of caching.""" def decorator(func): @wraps(func) def func_with_caching(*args, **kwargs): key = keyfunc(*args, **kwargs) val = cache_get(key, cache_name=cache_name) extra = "" if cache_name == 'database': extra = ".dbcache" if with_statsd_key is not None: metric_key = with_statsd_key else: metric_key = statsd_key(key) status = "hit" if val is not None else "miss" statsd.incr("cache%s.%s.%s" % (extra, metric_key, status)) # Values are singleton tuples so that we can distinguish # a result of None from a missing key. if val is not None: return val[0] val = func(*args, **kwargs) cache_set(key, val, cache_name=cache_name, timeout=timeout) return val return func_with_caching return decorator def cache_set(key, val, cache_name=None, timeout=None): memcached_stats_start() cache_backend = get_cache_backend(cache_name) ret = cache_backend.set(KEY_PREFIX + key, (val,), timeout=timeout) memcached_stats_finish() return ret def cache_get(key, cache_name=None): memcached_stats_start() cache_backend = get_cache_backend(cache_name) ret = cache_backend.get(KEY_PREFIX + key) memcached_stats_finish() return ret def cache_get_many(keys, cache_name=None): keys = [KEY_PREFIX + key for key in keys] memcached_stats_start() ret = get_cache_backend(cache_name).get_many(keys) memcached_stats_finish() return dict([(key[len(KEY_PREFIX):], value) for key, value in ret.items()]) def cache_set_many(items, cache_name=None, timeout=None): new_items = {} for key in items: new_items[KEY_PREFIX + key] = items[key] items = new_items memcached_stats_start() ret = get_cache_backend(cache_name).set_many(items, timeout=timeout) memcached_stats_finish() return ret def cache_delete(key, cache_name=None): memcached_stats_start() get_cache_backend(cache_name).delete(KEY_PREFIX + key) memcached_stats_finish() def cache_delete_many(items, cache_name=None): memcached_stats_start() get_cache_backend(cache_name).delete_many( KEY_PREFIX + item for item in items) memcached_stats_finish() # Required Arguments are as follows: # * object_ids: The list of object ids to look up # * cache_key_function: object_id => cache key # * query_function: [object_ids] => [objects from database] # Optional keyword arguments: # * setter: Function to call before storing items to cache (e.g. compression) # * extractor: Function to call on items returned from cache # (e.g. decompression). Should be the inverse of the setter # function. # * id_fetcher: Function mapping an object from database => object_id # (in case we're using a key more complex than obj.id) # * cache_transformer: Function mapping an object from database => # value for cache (in case the values that we're caching are some # function of the objects, not the objects themselves) def generic_bulk_cached_fetch(cache_key_function, query_function, object_ids, extractor=lambda obj: obj, setter=lambda obj: obj, id_fetcher=lambda obj: obj.id, cache_transformer=lambda obj: obj): cache_keys = {} for object_id in object_ids: cache_keys[object_id] = cache_key_function(object_id) cached_objects = cache_get_many([cache_keys[object_id] for object_id in object_ids]) for (key, val) in cached_objects.items(): cached_objects[key] = extractor(cached_objects[key][0]) needed_ids = [object_id for object_id in object_ids if cache_keys[object_id] not in cached_objects] db_objects = query_function(needed_ids) items_for_memcached = {} for obj in db_objects: key = cache_keys[id_fetcher(obj)] item = cache_transformer(obj) items_for_memcached[key] = (setter(item),) cached_objects[key] = item if len(items_for_memcached) > 0: cache_set_many(items_for_memcached) return dict((object_id, cached_objects[cache_keys[object_id]]) for object_id in object_ids if cache_keys[object_id] in cached_objects) def cache(func): """Decorator which applies Django caching to a function. Uses a key based on the function's name, filename, and the repr() of its arguments.""" func_uniqifier = '%s-%s' % (func.func_code.co_filename, func.func_name) @wraps(func) def keyfunc(*args, **kwargs): # Django complains about spaces because memcached rejects them key = func_uniqifier + repr((args, kwargs)) return key.replace('-','--').replace(' ','-s') return cache_with_key(keyfunc)(func) def message_cache_key(message_id): return "message:%d" % (message_id,) def display_recipient_cache_key(recipient_id): return "display_recipient_dict:%d" % (recipient_id,) def user_profile_by_email_cache_key(email): # See the comment in zerver/lib/avatar.py:gravatar_hash for why we # are proactively encoding email addresses even though they will # with high likelihood be ASCII-only for the foreseeable future. return 'user_profile_by_email:%s' % (make_safe_digest(email.strip()),) def user_profile_by_id_cache_key(user_profile_id): return "user_profile_by_id:%s" % (user_profile_id,) def cache_save_user_profile(user_profile): cache_set(user_profile_by_id_cache_key(user_profile.id), user_profile, timeout=3600*24*7) def active_user_dicts_in_realm_cache_key(realm): return "active_user_dicts_in_realm:%s" % (realm.id,) def get_stream_cache_key(stream_name, realm): from zerver.models import Realm if isinstance(realm, Realm): realm_id = realm.id else: realm_id = realm return "stream_by_realm_and_name:%s:%s" % ( realm_id, make_safe_digest(stream_name.strip().lower())) def update_user_profile_caches(user_profiles): items_for_memcached = {} for user_profile in user_profiles: items_for_memcached[user_profile_by_email_cache_key(user_profile.email)] = (user_profile,) items_for_memcached[user_profile_by_id_cache_key(user_profile.id)] = (user_profile,) cache_set_many(items_for_memcached) # Called by models.py to flush the user_profile cache whenever we save # a user_profile object def flush_user_profile(sender, **kwargs): user_profile = kwargs['instance'] update_user_profile_caches([user_profile]) # Invalidate our active_users_in_realm info dict if any user has changed # name or email if kwargs['update_fields'] is None or \ len(set(['full_name', 'short_name', 'email', 'is_active']) & set(kwargs['update_fields'])) > 0: cache_delete(active_user_dicts_in_realm_cache_key(user_profile.realm)) # Invalidate realm-wide alert words cache if any user in the realm has changed # alert words if kwargs['update_fields'] is None or "alert_words" in kwargs['update_fields']: cache_delete(realm_alert_words_cache_key(user_profile.realm)) # Called by models.py to flush various caches whenever we save # a Realm object. The main tricky thing here is that Realm info is # generally cached indirectly through user_profile objects. def flush_realm(sender, **kwargs): realm = kwargs['instance'] users = realm.get_active_users() update_user_profile_caches(users) if realm.deactivated: cache_delete(active_user_dicts_in_realm_cache_key(realm)) cache_delete(realm_alert_words_cache_key(realm)) def realm_alert_words_cache_key(realm): return "realm_alert_words:%s" % (realm.domain,) # Called by models.py to flush the stream cache whenever we save a stream # object. def flush_stream(sender, **kwargs): stream = kwargs['instance'] items_for_memcached = {} items_for_memcached[get_stream_cache_key(stream.name, stream.realm)] = (stream,) cache_set_many(items_for_memcached)