summaryrefslogtreecommitdiff
path: root/pymemcache/client/hash.py
blob: 9c3e68efa871accb2646e6d498e74e9dd407bff7 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
import socket
import time
import logging

from pymemcache.client.base import Client, PooledClient, _check_key
from pymemcache.client.rendezvous import RendezvousHash

logger = logging.getLogger(__name__)


class HashClient(object):
    """
    A client for communicating with a cluster of memcached servers
    """
    def __init__(
        self,
        servers,
        hasher=None,
        serializer=None,
        deserializer=None,
        connect_timeout=None,
        timeout=None,
        no_delay=False,
        socket_module=socket,
        key_prefix=b'',
        max_pool_size=None,
        lock_generator=None,
        retry_attempts=2,
        retry_timeout=1,
        dead_timeout=60,
        use_pooling=False,
        ignore_exc=False,
    ):
        """
        Constructor.

        Args:
          servers: list(tuple(hostname, port))
          hasher: optional class three functions ``get_node``, ``add_node``,
                  and ``remove_node``
                  defaults to Rendezvous (HRW) hash.

          use_pooling: use py:class:`.PooledClient` as the default underlying
                       class. ``max_pool_size`` and ``lock_generator`` can
                       be used with this. default: False

          retry_attempts: Amount of times a client should be tried before it
                          is marked dead and removed from the pool.
          retry_timeout (float): Time in seconds that should pass between retry
                                 attempts.
          dead_timeout (float): Time in seconds before attempting to add a node
                                back in the pool.

        Further arguments are interpreted as for :py:class:`.Client`
        constructor.

        The default ``hasher`` is using a pure python implementation that can
        be significantly improved performance wise by switching to a C based
        version. We recommend using ``python-clandestined`` if having a C
        dependency is acceptable.
        """
        self.clients = {}
        self.retry_attempts = retry_attempts
        self.retry_timeout = retry_timeout
        self.dead_timeout = dead_timeout
        self.use_pooling = use_pooling
        self.key_prefix = key_prefix
        self.ignore_exc = ignore_exc
        self._failed_clients = {}
        self._dead_clients = {}
        self._last_dead_check_time = time.time()

        if hasher is None:
            self.hasher = RendezvousHash()

        self.default_kwargs = {
            'connect_timeout': connect_timeout,
            'timeout': timeout,
            'no_delay': no_delay,
            'socket_module': socket_module,
            'key_prefix': key_prefix,
            'serializer': serializer,
            'deserializer': deserializer,
        }

        if use_pooling is True:
            self.default_kwargs.update({
                'max_pool_size': max_pool_size,
                'lock_generator': lock_generator
            })

        for server, port in servers:
            self.add_server(server, port)

    def add_server(self, server, port):
        key = '%s:%s' % (server, port)

        if self.use_pooling:
            client = PooledClient(
                (server, port),
                **self.default_kwargs
            )
        else:
            client = Client((server, port), **self.default_kwargs)

        self.clients[key] = client
        self.hasher.add_node(key)

    def remove_server(self, server, port):
        dead_time = time.time()
        self._failed_clients.pop((server, port))
        self._dead_clients[(server, port)] = dead_time
        key = '%s:%s' % (server, port)
        self.hasher.remove_node(key)

    def _get_client(self, key):
        _check_key(key, self.key_prefix)
        if len(self._dead_clients) > 0:
            current_time = time.time()
            ldc = self._last_dead_check_time
            # we have dead clients and we have reached the
            # timeout retry
            if current_time - ldc > self.dead_timeout:
                for server, dead_time in self._dead_clients.items():
                    if current_time - dead_time > self.dead_timeout:
                        logger.debug(
                            'bringing server back into rotation %s',
                            server
                        )
                        self.add_server(*server)
                        self._last_dead_check_time = current_time

        server = self.hasher.get_node(key)
        client = self.clients[server]
        return client

    def _safely_run_func(self, client, func, default_val, *args, **kwargs):
        try:
            if client.server in self._failed_clients:
                # This server is currently failing, lets check if it is in
                # retry or marked as dead
                failed_metadata = self._failed_clients[client.server]

                # we haven't tried our max amount yet, if it has been enough
                # time lets just retry using it
                if failed_metadata['attempts'] < self.retry_attempts:
                    failed_time = failed_metadata['failed_time']
                    if time.time() - failed_time > self.retry_timeout:
                        logger.debug(
                            'retrying failed server: %s', client.server
                        )
                        result = func(*args, **kwargs)
                        # we were successful, lets remove it from the failed
                        # clients
                        self._failed_clients.pop(client.server)
                        return result
                    return default_val
                else:
                    # We've reached our max retry attempts, we need to mark
                    # the sever as dead
                    logger.debug('marking server as dead: %s', client.server)
                    self.remove_server(*client.server)

            result = func(*args, **kwargs)
            return result

        # Connecting to the server fail, we should enter
        # retry mode
        except socket.error:
            # This client has never failed, lets mark it for failure
            if (
                    client.server not in self._failed_clients and
                    self.retry_attempts > 0
            ):
                self._failed_clients[client.server] = {
                    'failed_time': time.time(),
                    'attempts': 0,
                }
            # We aren't allowing any retries, we should mark the server as
            # dead immediately
            elif (
                client.server not in self._failed_clients and
                self.retry_attempts < 0
            ):
                self._failed_clients[client.server] = {
                    'failed_time': time.time(),
                    'attempts': 0,
                }
                logger.debug("marking server as dead %s" % client.server)
                self.remove_server(*client.server)
            # This client has failed previously, we need to update the metadata
            # to reflect that we have attempted it again
            else:
                failed_metadata = self._failed_clients[client.server]
                failed_metadata['attempts'] += 1
                failed_metadata['failed_time'] = time.time()
                self._failed_clients[client.server] = failed_metadata

            # if we haven't enabled ignore_exc, don't move on gracefully, just
            # raise the exception
            if not self.ignore_exc:
                raise

            return default_val
        except:
            # any exceptions that aren't socket.error we need to handle
            # gracefully as well
            if not self.ignore_exc:
                raise

            return default_val

    def _run_cmd(self, cmd, key, default_val, *args, **kwargs):
        client = self._get_client(key)
        func = getattr(client, cmd)
        args = list(args)
        args.insert(0, key)
        return self._safely_run_func(
            client, func, default_val, *args, **kwargs
        )

    def set(self, key, *args, **kwargs):
        return self._run_cmd('set', key, False, *args, **kwargs)

    def get(self, key, *args, **kwargs):
        return self._run_cmd('get', key, None, *args, **kwargs)

    def incr(self, key, *args, **kwargs):
        return self._run_cmd('incr', key, False, *args, **kwargs)

    def decr(self, key, *args, **kwargs):
        return self._run_cmd('decr', key, False, *args, **kwargs)

    def set_many(self, values, *args, **kwargs):
        client_batches = {}
        for key, value in values.items():
            client = self._get_client(key)
            if client.server not in client_batches:
                client_batches[client.server] = {}

            client_batches[client.server][key] = value

        end = []

        for server, values in client_batches.items():
            client = self.clients['%s:%s' % server]
            new_args = list(args)
            new_args.insert(0, values)
            result = self._safely_run_func(
                client,
                client.set_many, False, *new_args, **kwargs
            )
            end.append(result)

        return all(end)

    set_multi = set_many

    def get_many(self, keys, *args, **kwargs):
        client_batches = {}
        for key in keys:
            client = self._get_client(key)
            if client.server not in client_batches:
                client_batches[client.server] = []

            client_batches[client.server].append(key)

        end = {}

        for server, keys in client_batches.items():
            client = self.clients['%s:%s' % server]
            new_args = list(args)
            new_args.insert(0, keys)
            result = self._safely_run_func(
                client,
                client.get_many, {}, *new_args, **kwargs
            )
            end.update(result)

        return end

    get_multi = get_many

    def gets(self, key, *args, **kwargs):
        return self._run_cmd('gets', key, None, *args, **kwargs)

    def add(self, key, *args, **kwargs):
        return self._run_cmd('add', key, False, *args, **kwargs)

    def prepend(self, key, *args, **kwargs):
        return self._run_cmd('prepend', key, False, *args, **kwargs)

    def append(self, key, *args, **kwargs):
        return self._run_cmd('append', key, False, *args, **kwargs)

    def delete(self, key, *args, **kwargs):
        return self._run_cmd('delete', key, False, *args, **kwargs)

    def cas(self, key, *args, **kwargs):
        return self._run_cmd('cas', key, False, *args, **kwargs)

    def replace(self, key, *args, **kwargs):
        return self._run_cmd('replace', key, False, *args, **kwargs)

    def flush_all(self):
        for _, client in self.clients.items():
            self._safely_run_func(client, client.flush_all, False)