IGNITE-14851 Enable partition awareness by default, fix unnecessary double connect - Fixes #42.

Author: ivandasch

docs/async_examples.rst

@@ -84,6 +84,7 @@ File: `transactions.py`_.
 
 Client transactions are supported for caches with
 :py:attr:`~pyignite.datatypes.cache_config.CacheAtomicityMode.TRANSACTIONAL` mode.
+**Supported only python 3.7+**
 
 Let's create transactional cache:
 
@@ -97,28 +98,28 @@ Let's start a transaction and commit it:
 .. literalinclude:: ../examples/transactions.py
   :language: python
   :dedent: 8
-  :lines: 36-41
+  :lines: 36-42
 
 Let's check that the transaction was committed successfully:
 
 .. literalinclude:: ../examples/transactions.py
   :language: python
   :dedent: 8
-  :lines: 44-45
+  :lines: 45-46
 
 Let's check that raising exception inside `async with` block leads to transaction's rollback
 
 .. literalinclude:: ../examples/transactions.py
   :language: python
   :dedent: 8
-  :lines: 48-59
+  :lines: 49-61
 
 Let's check that timed out transaction is successfully rolled back
 
 .. literalinclude:: ../examples/transactions.py
   :language: python
   :dedent: 8
-  :lines: 62-73
+  :lines: 64-75
 
 See more info about transaction's parameters in a documentation of :py:meth:`~pyignite.aio_client.AioClient.tx_start`
 

docs/examples.rst

@@ -219,35 +219,35 @@ Let's create transactional cache:
 .. literalinclude:: ../examples/transactions.py
   :language: python
   :dedent: 8
-  :lines: 82-85
+  :lines: 84-87
 
 Let's start a transaction and commit it:
 
 .. literalinclude:: ../examples/transactions.py
   :language: python
   :dedent: 8
-  :lines: 88-92
+  :lines: 90-96
 
 Let's check that the transaction was committed successfully:
 
 .. literalinclude:: ../examples/transactions.py
   :language: python
   :dedent: 8
-  :lines: 94-95
+  :lines: 98-99
 
 Let's check that raising exception inside `with` block leads to transaction's rollback
 
 .. literalinclude:: ../examples/transactions.py
   :language: python
   :dedent: 8
-  :lines: 98-108
+  :lines: 102-113
 
 Let's check that timed out transaction is successfully rolled back
 
 .. literalinclude:: ../examples/transactions.py
   :language: python
   :dedent: 8
-  :lines: 111-121
+  :lines: 116-126
 
 See more info about transaction's parameters in a documentation of :py:meth:`~pyignite.client.Client.tx_start`
 

docs/images/partitionawareness01.png

@@ -22,6 +22,7 @@ Welcome to Apache Ignite binary client Python API documentation!
 
    readme
    modules
+   partition_awareness
    examples
    async_examples
 

docs/images/partitionawareness02.png

@@ -0,0 +1,63 @@
+..  Licensed to the Apache Software Foundation (ASF) under one or more
+    contributor license agreements.  See the NOTICE file distributed with
+    this work for additional information regarding copyright ownership.
+    The ASF licenses this file to You under the Apache License, Version 2.0
+    (the "License"); you may not use this file except in compliance with
+    the License.  You may obtain a copy of the License at
+
+..      http://www.apache.org/licenses/LICENSE-2.0
+
+..  Unless required by applicable law or agreed to in writing, software
+    distributed under the License is distributed on an "AS IS" BASIS,
+    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+    See the License for the specific language governing permissions and
+    limitations under the License.
+
+===================
+Partition Awareness
+===================
+
+Partition awareness allows the thin client to send query requests directly to the node that owns the queried data.
+
+Without partition awareness, an application that is connected to the cluster via a thin client executes all queries and operations via a single server node that acts as a proxy for the incoming requests. These operations are then re-routed to the node that stores the data that is being requested. This results in a bottleneck that could prevent the application from scaling linearly.
+
+.. image:: images/partitionawareness01.png
+  :alt: Without partition awareness
+
+Notice how queries must pass through the proxy server node, where they are routed to the correct node.
+
+With partition awareness in place, the thin client can directly route queries and operations to the primary nodes that own the data required for the queries. This eliminates the bottleneck, allowing the application to scale more easily.
+
+.. image:: images/partitionawareness02.png
+  :alt: With partition awareness
+
+Partition awareness can be enabled or disabled by setting `partition_aware` parameter in
+:meth:`pyignite.client.Client.__init__` or :meth:`pyignite.aio_client.AioClient.__init__` to `True` (by default)
+or `False`.
+
+Also, it is recommended to pass list of address and port pairs of all server nodes
+to :meth:`pyignite.client.Client.connect` or to :meth:`pyignite.aio_client.AioClient.connect`.
+
+For example:
+
+.. code-block:: python3
+
+    from pyignite import Client
+
+    client = Client(
+        partition_awareness=True
+    )
+    nodes = [('10.128.0.1', 10800), ('10.128.0.2', 10800),...]
+    with client.connect(nodes):
+        ....
+
+.. code-block:: python3
+
+    from pyignite import AioClient
+
+    client = AioClient(
+        partition_awareness=True
+    )
+    nodes = [('10.128.0.1', 10800), ('10.128.0.2', 10800),...]
+    async with client.connect(nodes):
+        ....

docs/index.rst

@@ -14,7 +14,7 @@
     limitations under the License.
 
 pyignite.transaction module
-=========================
+===========================
 
 .. automodule:: pyignite.transaction
     :members:

docs/partition_awareness.rst

@@ -35,19 +35,21 @@ async def async_example():
         # starting transaction
         key = 1
         async with client.tx_start(
-                isolation=TransactionIsolation.REPEATABLE_READ, concurrency=TransactionConcurrency.PESSIMISTIC
+                isolation=TransactionIsolation.REPEATABLE_READ,
+                concurrency=TransactionConcurrency.PESSIMISTIC
         ) as tx:
             await cache.put(key, 'success')
             await tx.commit()
 
         # key=1 value=success
         val = await cache.get(key)
-        print(f"key=1 value={val}")
+        print(f"key={key} value={val}")
 
         # rollback transaction.
         try:
             async with client.tx_start(
-                    isolation=TransactionIsolation.REPEATABLE_READ, concurrency=TransactionConcurrency.PESSIMISTIC
+                    isolation=TransactionIsolation.REPEATABLE_READ,
+                    concurrency=TransactionConcurrency.PESSIMISTIC
             ):
                 await cache.put(key, 'fail')
                 raise RuntimeError('test')
@@ -56,7 +58,7 @@ async def async_example():
 
         # key=1 value=success
         val = await cache.get(key)
-        print(f"key=1 value={val}")
+        print(f"key={key} value={val}")
 
         # rollback transaction on timeout.
         try:
@@ -70,7 +72,7 @@ async def async_example():
 
         # key=1 value=success
         val = await cache.get(key)
-        print(f"key=1 value={val}")
+        print(f"key={key} value={val}")
 
         # destroy cache
         await cache.destroy()
@@ -85,40 +87,43 @@ def sync_example():
         })
 
         # starting transaction
+        key = 1
         with client.tx_start(
-                isolation=TransactionIsolation.REPEATABLE_READ, concurrency=TransactionConcurrency.PESSIMISTIC
+                isolation=TransactionIsolation.REPEATABLE_READ,
+                concurrency=TransactionConcurrency.PESSIMISTIC
         ) as tx:
-            cache.put(1, 'success')
+            cache.put(key, 'success')
             tx.commit()
 
         # key=1 value=success
-        print(f"key=1 value={cache.get(1)}")
+        print(f"key={key} value={cache.get(key)}")
 
         # rollback transaction.
         try:
             with client.tx_start(
-                    isolation=TransactionIsolation.REPEATABLE_READ, concurrency=TransactionConcurrency.PESSIMISTIC
+                    isolation=TransactionIsolation.REPEATABLE_READ,
+                    concurrency=TransactionConcurrency.PESSIMISTIC
             ):
-                cache.put(1, 'fail')
+                cache.put(key, 'fail')
                 raise RuntimeError('test')
         except RuntimeError:
             pass
 
         # key=1 value=success
-        print(f"key=1 value={cache.get(1)}")
+        print(f"key={key} value={cache.get(key)}")
 
         # rollback transaction on timeout.
         try:
             with client.tx_start(timeout=1.0, label='long-tx') as tx:
-                cache.put(1, 'fail')
+                cache.put(key, 'fail')
                 time.sleep(2.0)
                 tx.commit()
         except CacheError as e:
             # Cache transaction timed out: GridNearTxLocal[...timeout=1000, ... label=long-tx]
             print(e)
 
         # key=1 value=success
-        print(f"key=1 value={cache.get(1)}")
+        print(f"key={key} value={cache.get(key)}")
 
         # destroy cache
         cache.destroy()

docs/source/pyignite.transaction.rst

@@ -60,21 +60,18 @@ class AioClient(BaseClient):
     Asynchronous Client implementation.
     """
 
-    def __init__(self, compact_footer: bool = None, partition_aware: bool = False, **kwargs):
+    def __init__(self, compact_footer: bool = None, partition_aware: bool = True, **kwargs):
         """
         Initialize client.
 
         :param compact_footer: (optional) use compact (True, recommended) or
          full (False) schema approach when serializing Complex objects.
          Default is to use the same approach the server is using (None).
          Apache Ignite binary protocol documentation on this topic:
-         https://apacheignite.readme.io/docs/binary-client-protocol-data-format#section-schema
+         https://ignite.apache.org/docs/latest/binary-client-protocol/data-format#schema
         :param partition_aware: (optional) try to calculate the exact data
          placement from the key before to issue the key operation to the
-         server node:
-         https://cwiki.apache.org/confluence/display/IGNITE/IEP-23%3A+Best+Effort+Affinity+for+thin+clients
-         The feature is in experimental status, so the parameter is `False`
-         by default. This will be changed later.
+         server node, `True` by default.
         """
         super().__init__(compact_footer, partition_aware, **kwargs)
         self._registry_mux = asyncio.Lock()
@@ -494,7 +491,7 @@ def tx_start(self, concurrency: TransactionConcurrency = TransactionConcurrency.
                  isolation: TransactionIsolation = TransactionIsolation.REPEATABLE_READ,
                  timeout: Union[int, float] = 0, label: Optional[str] = None) -> 'AioTransaction':
         """
-        Start async thin client transaction.
+        Start async thin client transaction. **Supported only python 3.7+**
 
         :param concurrency: (optional) transaction concurrency, see
                 :py:class:`~pyignite.datatypes.transactions.TransactionConcurrency`

examples/transactions.py

@@ -17,7 +17,7 @@
 This module contains functions, that are (more or less) directly mapped to
 Apache Ignite binary protocol operations. Read more:
 
-https://apacheignite.readme.io/docs/binary-client-protocol#section-client-operations
+https://ignite.apache.org/docs/latest/binary-client-protocol/binary-client-protocol#client-operations
 
 When the binary client protocol changes, these functions also change. For
 stable end user API see :mod:`pyignite.client` module.