@@ -9,192 +9,18 @@ Retryable Reads Tests
9
9
Introduction
10
10
============
11
11
12
- The YAML and JSON files in the ``legacy `` and ``unified `` sub-directories are platform-independent tests
13
- that drivers can use to prove their conformance to the Retryable Reads spec. Tests in the
14
- ``unified `` directory are written using the `Unified Test Format <../../unified-test-format/unified-test-format.md >`_.
15
- Tests in the ``legacy `` directory are written using the format described below.
12
+ The YAML and JSON files in this directory are platform-independent tests meant
13
+ to exercise a driver's implementation of retryable reads. These tests utilize
14
+ the [Unified Test Format](../../unified-test-format/unified-test-format.md).
16
15
17
- Prose tests, which are not easily expressed in YAML, are also presented
16
+ Several prose tests, which are not easily expressed in YAML, are also presented
18
17
in this file. Those tests will need to be manually implemented by each driver.
19
18
20
- Tests will require a MongoClient created with options defined in the tests.
21
- Integration tests will require a running MongoDB cluster with server versions
22
- 4.0 or later.
19
+ Prose Tests
20
+ ===========
23
21
24
- N.B. The spec specifies 3.6 as the minimum server version: however,
25
- ``failCommand `` is not supported on 3.6, so for now, testing requires MongoDB
26
- 4.0. Once `DRIVERS-560 `_ is resolved, we will attempt to adapt its live failure
27
- integration tests to test Retryable Reads on MongoDB 3.6.
28
-
29
- .. _DRIVERS-560 : https://jira.mongodb.org/browse/DRIVERS-560
30
-
31
- Server Fail Point
32
- =================
33
-
34
- See: `Server Fail Point `_ in the Transactions spec test suite.
35
-
36
- .. _Server Fail Point : ../../transactions/tests#server-fail-point
37
-
38
- Disabling Fail Point after Test Execution
39
- -----------------------------------------
40
-
41
- After each test that configures a fail point, drivers should disable the
42
- ``failCommand `` fail point to avoid spurious failures in
43
- subsequent tests. The fail point may be disabled like so::
44
-
45
- db.runCommand({
46
- configureFailPoint: "failCommand",
47
- mode: "off"
48
- });
49
-
50
- Network Error Tests
51
- ===================
52
-
53
- Network error tests are expressed in YAML and should be run against a standalone,
54
- shard cluster, or single-node replica set.
55
-
56
-
57
- Test Format
58
- -----------
59
-
60
- Each YAML file has the following keys:
61
-
62
- - ``runOn `` (optional): An array of server version and/or topology requirements
63
- for which the tests can be run. If the test environment satisfies one or more
64
- of these requirements, the tests may be executed; otherwise, this file should
65
- be skipped. If this field is omitted, the tests can be assumed to have no
66
- particular requirements and should be executed. Each element will have some or
67
- all of the following fields:
68
-
69
- - ``minServerVersion `` (optional): The minimum server version (inclusive)
70
- required to successfully run the tests. If this field is omitted, it should
71
- be assumed that there is no lower bound on the required server version.
72
-
73
- - ``maxServerVersion `` (optional): The maximum server version (inclusive)
74
- against which the tests can be run successfully. If this field is omitted,
75
- it should be assumed that there is no upper bound on the required server
76
- version.
77
-
78
- - ``topology `` (optional): An array of server topologies against which the
79
- tests can be run successfully. Valid topologies are "single",
80
- "replicaset", "sharded", and "load-balanced". If this field is omitted,
81
- the default is all topologies (i.e. ``["single", "replicaset", "sharded",
82
- "load-balanced"] ``).
83
-
84
- - ``serverless ``: (optional): Whether or not the test should be run on Atlas
85
- Serverless instances. Valid values are "require", "forbid", and "allow". If
86
- "require", the test MUST only be run on Atlas Serverless instances. If
87
- "forbid", the test MUST NOT be run on Atlas Serverless instances. If omitted
88
- or "allow", this option has no effect.
89
-
90
- The test runner MUST be informed whether or not Atlas Serverless is being
91
- used in order to determine if this requirement is met (e.g. through an
92
- environment variable or configuration option).
93
-
94
- Note: the Atlas Serverless proxy imitates mongos, so the test runner is not
95
- capable of determining if Atlas Serverless is in use by issuing commands
96
- such as ``buildInfo `` or ``hello ``. Furthermore, connections to Atlas
97
- Serverless use a load balancer, so the topology will appear as
98
- "load-balanced".
99
-
100
- - ``database_name `` and ``collection_name ``: Optional. The database and
101
- collection to use for testing.
102
-
103
- - ``bucket_name ``: Optional. The GridFS bucket name to use for testing.
104
-
105
- - ``data ``: The data that should exist in the collection(s) under test before
106
- each test run. This will typically be an array of documents to be inserted
107
- into the collection under test (i.e. ``collection_name ``); however, this field
108
- may also be an object mapping collection names to arrays of documents to be
109
- inserted into the specified collection.
110
-
111
- - ``tests ``: An array of tests that are to be run independently of each other.
112
- Each test will have some or all of the following fields:
113
-
114
- - ``description ``: The name of the test.
115
-
116
- - ``clientOptions ``: Optional, parameters to pass to MongoClient().
117
-
118
- - ``useMultipleMongoses `` (optional): If ``true ``, and the topology type is
119
- ``Sharded ``, the MongoClient for this test should be initialized with multiple
120
- mongos seed addresses. If ``false `` or omitted, only a single mongos address
121
- should be specified.
122
-
123
- If ``true ``, the topology type is ``LoadBalanced ``, and Atlas Serverless is
124
- not being used, the MongoClient for this test should be initialized with the
125
- URI of the load balancer fronting multiple servers. If ``false `` or omitted,
126
- the MongoClient for this test should be initialized with the URI of the load
127
- balancer fronting a single server.
128
-
129
- ``useMultipleMongoses `` only affects ``Sharded `` and ``LoadBalanced ``
130
- topologies (excluding Atlas Serverless).
131
-
132
- - ``skipReason ``: Optional, string describing why this test should be skipped.
133
-
134
- - ``failPoint ``: Optional, a server fail point to enable, expressed as the
135
- configureFailPoint command to run on the admin database.
136
-
137
- - ``operations ``: An array of documents describing an operation to be
138
- executed. Each document has the following fields:
139
-
140
- - ``name ``: The name of the operation on ``object ``.
141
-
142
- - ``object ``: The name of the object to perform the operation on. Can be
143
- "database", "collection", "client", or "gridfsbucket."
144
-
145
- - ``arguments ``: Optional, the names and values of arguments.
146
-
147
- - ``result ``: Optional. The return value from the operation, if any. This
148
- field may be a scalar (e.g. in the case of a count), a single document, or
149
- an array of documents in the case of a multi-document read.
150
-
151
- - ``error ``: Optional. If ``true ``, the test should expect an error or
152
- exception.
153
-
154
- - ``expectations ``: Optional list of command-started events.
155
-
156
- GridFS Tests
157
- ------------
158
-
159
- GridFS tests are denoted by when the YAML file contains ``bucket_name ``.
160
- The ``data `` field will also be an object, which maps collection names
161
- (e.g. ``fs.files ``) to an array of documents that should be inserted into
162
- the specified collection.
163
-
164
- ``fs.files `` and ``fs.chunks `` should be created in the database
165
- specified by ``database_name ``. This could be done via inserts or by
166
- creating GridFSBuckets—using the GridFS ``bucketName `` (see
167
- `GridFSBucket spec `_) specified by ``bucket_name `` field in the YAML
168
- file—and calling ``upload_from_stream_with_id `` with the appropriate
169
- data.
170
-
171
- ``Download `` tests should be tested against ``GridFS.download_to_stream ``.
172
- ``DownloadByName `` tests should be tested against
173
- ``GridFS.download_to_stream_by_name ``.
174
-
175
-
176
- .. _GridFSBucket spec : https://github.com/mongodb/specifications/blob/master../../gridfs/gridfs-spec.md#configurable-gridfsbucket-class
177
-
178
-
179
- Speeding Up Tests
180
- -----------------
181
-
182
- Drivers can greatly reduce the execution time of tests by setting `heartbeatFrequencyMS `_
183
- and `minHeartbeatFrequencyMS `_ (internally) to a small value (e.g. 5ms), below what
184
- is normally permitted in the SDAM spec. If a test specifies an explicit value for
185
- heartbeatFrequencyMS (e.g. client or URI options), drivers MUST use that value.
186
-
187
- .. _minHeartbeatFrequencyMS : ../../server-discovery-and-monitoring/server-discovery-and-monitoring.rst#minheartbeatfrequencyms
188
- .. _heartbeatFrequencyMS : ../../server-discovery-and-monitoring/server-discovery-and-monitoring.rst#heartbeatfrequencyms
189
-
190
- Optional Enumeration Commands
191
- =============================
192
-
193
- A driver only needs to test the optional enumeration commands it has chosen to
194
- implement (e.g. ``Database.listCollectionNames() ``).
195
-
196
- PoolClearedError Retryability Test
197
- ==================================
22
+ 1. PoolClearedError Retryability Test
23
+ -------------------------------------
198
24
199
25
This test will be used to ensure drivers properly retry after encountering PoolClearedErrors.
200
26
It MUST be implemented by any driver that implements the CMAP specification.
@@ -232,8 +58,8 @@ This test requires MongoDB 4.2.9+ for ``blockConnection`` support in the failpoi
232
58
233
59
9. Disable the failpoint.
234
60
235
- Retrying Reads in a Sharded Cluster
236
- ===================================
61
+ 2. Retrying Reads in a Sharded Cluster
62
+ --------------------------------------
237
63
238
64
These tests will be used to ensure drivers properly retry reads on a different
239
65
mongos.
@@ -244,8 +70,8 @@ different mongos due to normal SDAM behavior of randomized suitable server
244
70
selection". Verify relevant code paths are correctly executed by the tests using
245
71
external means such as a logging, debugger, code coverage tool, etc.
246
72
247
- Retryable Reads Are Retried on a Different mongos When One is Available
248
- -----------------------------------------------------------------------
73
+ 2.1 Retryable Reads Are Retried on a Different mongos When One is Available
74
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
249
75
250
76
This test MUST be executed against a sharded cluster that has at least two
251
77
mongos instances, supports ``retryReads=true ``, and has enabled the
@@ -278,8 +104,8 @@ mongos instances, supports ``retryReads=true``, and has enabled the
278
104
7. Disable the fail point on both ``s0 `` and ``s1 ``.
279
105
280
106
281
- Retryable Reads Are Retried on the Same mongos When No Others are Available
282
- ---------------------------------------------------------------------------
107
+ 2.2 Retryable Reads Are Retried on the Same mongos When No Others are Available
108
+ ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
283
109
284
110
This test MUST be executed against a sharded cluster that supports
285
111
``retryReads=true `` and has enabled the ``configureFailPoint `` command
@@ -316,10 +142,12 @@ This test MUST be executed against a sharded cluster that supports
316
142
Changelog
317
143
=========
318
144
145
+ :2024-03-06: Convert legacy retryable reads tests to unified format.
146
+
319
147
:2024-02-21: Update mongos redirection prose tests to workaround SDAM behavior
320
148
preventing execution of deprioritization code paths.
321
149
322
- :2023-08-26 Add prose tests for retrying in a sharded cluster.
150
+ :2023-08-26: Add prose tests for retrying in a sharded cluster.
323
151
324
152
:2022-04-22: Clarifications to ``serverless `` and ``useMultipleMongoses ``.
325
153
0 commit comments