WIP on the re-implementation of fact caching and various backends.

lib/ansible/cache/__init__.py

@@ -0,0 +1,59 @@
+# (c) 2014, Michael DeHaan <michael.dehaan@gmail.com>
+#
+# This file is part of Ansible
+#
+# Ansible is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# Ansible is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with Ansible.  If not, see <http://www.gnu.org/licenses/>.
+
+from collections import MutableMapping
+
+from ansible import utils
+from ansible import constants as C
+from ansible import errors
+
+
+class FactCache(MutableMapping):
+
+    def __init__(self, *args, **kwargs):
+        self._plugin = utils.plugins.cache_loader.get(C.CACHE_PLUGIN)
+        if self._plugin is None:
+            return
+
+    def __getitem__(self, key):
+        if key not in self:
+            raise KeyError
+        return self._plugin.get(key)
+
+    def __setitem__(self, key, value):
+        self._plugin.set(key, value)
+
+    def __delitem__(self, key):
+        self._plugin.delete(key)
+
+    def __contains__(self, key):
+        return self._plugin.contains(key)
+
+    def __iter__(self):
+        return iter(self._plugin.keys())
+
+    def __len__(self):
+        return len(self._plugin.keys())
+
+    def copy(self):
+        """
+        Return a primitive copy of the keys and values from the cache.
+        """
+        return dict([(k, v) for (k, v) in self.iteritems()])
+
+    def keys(self):
+        return self._plugin.keys()

lib/ansible/cache/base.py

@@ -0,0 +1,15 @@
+class BaseCacheModule(object):
+    def get(self, key):
+        raise NotImplementedError("Subclasses of {} must implement the '{}' method".format(self.__class__.__name__, self.__name__))
+
+    def set(self, key, value):
+        raise NotImplementedError("Subclasses of {} must implement the '{}' method".format(self.__class__.__name__, self.__name__))
+
+    def keys(self):
+        raise NotImplementedError("Subclasses of {} must implement the '{}' method".format(self.__class__.__name__, self.__name__))
+
+    def contains(self, key):
+        raise NotImplementedError("Subclasses of {} must implement the '{}' method".format(self.__class__.__name__, self.__name__))
+
+    def delete(self, key):
+        raise NotImplementedError("Subclasses of {} must implement the '{}' method".format(self.__class__.__name__, self.__name__))

lib/ansible/cache/file.py

@@ -0,0 +1,65 @@
+# (c) 2014, Michael DeHaan <michael.dehaan@gmail.com>
+#
+# This file is part of Ansible
+#
+# Ansible is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# Ansible is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with Ansible.  If not, see <http://www.gnu.org/licenses/>.
+from __future__ import absolute_import
+
+import collections
+import json
+import os
+import shutil
+import sys
+import tempfile
+from datetime import datetime
+
+from ansible import constants as C
+from ansible.cache.memory import CacheModule as MemoryCacheModule
+
+
+class CacheModule(MemoryCacheModule):
+    def __init__(self):
+        super(CacheModule, self).__init__()
+        self._timeout = int(C.CACHE_PLUGIN_TIMEOUT)
+        self._filename = '/tmp/ansible_facts.json'
+
+        if os.access(self._filename, os.R_OK):
+            mtime = datetime.fromtimestamp(os.path.getmtime(self._filename))
+            if self._timeout == 0 or (datetime.now() - mtime).total_seconds() < self._timeout:
+                with open(self._filename, 'rb') as f:
+                    # we could make assumptions about the MemoryCacheModule here if we wanted
+                    # to be more efficient, but performance isn't the priority with this module
+                    data = json.load(f)
+                    if isinstance(data, collections.Mapping):
+                        for k, v in data.items():
+                            super(CacheModule, self).set(k, v)
+
+    def set(self, *args, **kwargs):
+        super(CacheModule, self).set(*args, **kwargs)
+        self.flush()
+
+    def delete(self, *args, **kwargs):
+        super(CacheModule, self).delete(*args, **kwargs)
+        self.flush()
+
+    def flush(self):
+        temp = tempfile.TemporaryFile('r+b')
+
+        try:
+            json.dump(self._cache, temp, separators=(',', ':'))
+            temp.seek(0)
+            with open(self._filename, 'w+b') as f:
+                shutil.copyfileobj(temp, f)
+        finally:
+            temp.close()

lib/ansible/cache/memcached.py

@@ -0,0 +1,112 @@
+# (c) 2014, Michael DeHaan <michael.dehaan@gmail.com>
+#
+# This file is part of Ansible
+#
+# Ansible is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# Ansible is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with Ansible.  If not, see <http://www.gnu.org/licenses/>.
+import collections
+import sys
+import time
+
+from ansible import constants as C
+from ansible.cache.base import BaseCacheModule
+
+try:
+    import memcache
+except ImportError:
+    print 'python-memcached is required for the memcached fact cache'
+    sys.exit(1)
+
+
+class CacheModuleKeys(collections.MutableSet):
+    """
+    A set subclass that keeps track of insertion time and persists
+    the set in memcached.
+    """
+    PREFIX = 'ansible_cache_keys'
+
+    def __init__(self, cache, *args, **kwargs):
+        self._cache = cache
+        self._keyset = dict(*args, **kwargs)
+
+    def __contains__(self, key):
+        return key in self._keyset
+
+    def __iter__(self):
+        return iter(self._keyset)
+
+    def __len__(self):
+        return len(self._keyset)
+
+    def add(self, key):
+        self._keyset[key] = time.time()
+        self._cache.set(self.PREFIX, self._keyset)
+
+    def discard(self, key):
+        del self._keyset[key]
+        self._cache.set(self.PREFIX, self._keyset)
+
+    def remove_by_timerange(self, s_min, s_max):
+        for k in self._keyset.keys():
+            t = self._keyset[k]
+            if s_min < t < s_max:
+                del self._keyset[k]
+        self._cache.set(self.PREFIX, self._keyset)
+
+
+class CacheModule(BaseCacheModule):
+
+    def __init__(self, *args, **kwargs):
+        if C.CACHE_PLUGIN_CONNECTION:
+            connection = C.CACHE_PLUGIN_CONNECTION.split(',')
+        else:
+            connection = ['127.0.0.1:11211']
+
+        self._timeout = C.CACHE_PLUGIN_TIMEOUT
+        self._prefix = C.CACHE_PLUGIN_PREFIX
+        self._cache = memcache.Client(connection, debug=0)
+        self._keys = CacheModuleKeys(self._cache, self._cache.get(CacheModuleKeys.PREFIX) or [])
+
+    def _make_key(self, key):
+        return "{}{}".format(self._prefix, key)
+
+    def _expire_keys(self):
+        if self._timeout > 0:
+            expiry_age = time.time() - self._timeout
+        self._keys.remove_by_timerange(0, expiry_age)
+
+    def get(self, key):
+        value = self._cache.get(self._make_key(key))
+        # guard against the key not being removed from the zset;
+        # this could happen in cases where the timeout value is changed
+        # between invocations
+        if value is None:
+            self.delete(key)
+            raise KeyError
+        return value
+
+    def set(self, key, value):
+        self._cache.set(self._make_key(key), value, time=self._timeout)
+        self._keys.add(key)
+
+    def keys(self):
+        self._expire_keys()
+        return list(iter(self._keys))
+
+    def contains(self, key):
+        self._expire_keys()
+        return key in self._keys
+
+    def delete(self, key):
+        self._cache.delete(self._make_key(key))
+        self._keys.discard(key)

lib/ansible/cache/memory.py

@@ -0,0 +1,37 @@
+# (c) 2014, Michael DeHaan <michael.dehaan@gmail.com>
+#
+# This file is part of Ansible
+#
+# Ansible is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# Ansible is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with Ansible.  If not, see <http://www.gnu.org/licenses/>.
+
+
+class CacheModule(object):
+
+    def __init__(self, *args, **kwargs):
+        self._cache = {}
+
+    def get(self, key):
+        return self._cache.get(key)
+
+    def set(self, key, value):
+        self._cache[key] = value
+
+    def keys(self):
+        return self._cache.keys()
+
+    def contains(self, key):
+        return key in self._cache
+
+    def delete(self, key):
+        del self._cache[key]

lib/ansible/cache/redis.py

@@ -0,0 +1,108 @@
+# (c) 2014, Michael DeHaan <michael.dehaan@gmail.com>
+#
+# This file is part of Ansible
+#
+# Ansible is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# Ansible is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with Ansible.  If not, see <http://www.gnu.org/licenses/>.
+from __future__ import absolute_import
+
+import collections
+import pickle
+import sys
+import time
+
+from ansible import constants as C
+from ansible.cache.base import BaseCacheModule
+
+try:
+    from redis import StrictRedis
+except ImportError:
+    print "The 'redis' Python module is required for the redis fact cache"
+    sys.exit(1)
+
+
+class PickledRedis(StrictRedis):
+    """
+    A subclass of StricRedis that uses the pickle module to store and load
+    representations of the provided values.
+    """
+    def get(self, name):
+        pickled_value = super(PickledRedis, self).get(name)
+        if pickled_value is None:
+            return None
+        return pickle.loads(pickled_value)
+
+    def set(self, name, value, *args, **kwargs):
+        return super(PickledRedis, self).set(name, pickle.dumps(value), *args, **kwargs)
+
+    def setex(self, name, time, value):
+        return super(PickledRedis, self).setex(name, time, pickle.dumps(value))
+
+
+class CacheModule(BaseCacheModule):
+    """
+    A caching module backed by redis.
+
+    Keys are maintained in a zset with their score being the timestamp
+    when they are inserted. This allows for the usage of 'zremrangebyscore'
+    to expire keys. This mechanism is used or a pattern matched 'scan' for
+    performance.
+    """
+    def __init__(self, *args, **kwargs):
+        if C.CACHE_PLUGIN_CONNECTION:
+            connection = C.CACHE_PLUGIN_CONNECTION.split(':')
+        else:
+            connection = []
+
+        self._timeout = C.CACHE_PLUGIN_TIMEOUT
+        self._prefix = C.CACHE_PLUGIN_PREFIX
+        self._cache = PickledRedis(*connection)
+        self._keys_set = 'ansible_cache_keys'
+
+    def _make_key(self, key):
+        return "{}{}".format(self._prefix, key)
+
+    def get(self, key):
+        value = self._cache.get(self._make_key(key))
+        # guard against the key not being removed from the zset;
+        # this could happen in cases where the timeout value is changed
+        # between invocations
+        if value is None:
+            self.delete(key)
+            raise KeyError
+        return value
+
+    def set(self, key, value):
+        if self._timeout > 0: # a timeout of 0 is handled as meaning 'never expire'
+            self._cache.setex(self._make_key(key), self._timeout, value)
+        else:
+            self._cache.set(self._make_key(key), value)
+
+        self._cache.zadd(self._keys_set, time.time(), key)
+
+    def _expire_keys(self):
+        if self._timeout > 0:
+            expiry_age = time.time() - self._timeout
+        self._cache.zremrangebyscore(self._keys_set, 0, expiry_age)
+
+    def keys(self):
+        self._expire_keys()
+        return self._cache.zrange(self._keys_set, 0, -1)
+
+    def contains(self, key):
+        self._expire_keys()
+        return (self._cache.zrank(self._keys_set, key) >= 0)
+
+    def delete(self, key):
+        self._cache.delete(self._make_key(key))
+        self._cache.zrem(self._keys_set, key)