|
| 1 | +Tarantool 3.1 |
| 2 | +============= |
| 3 | + |
| 4 | +Planned release date: April 2024 |
| 5 | + |
| 6 | +Releases on GitHub: not released yet |
| 7 | + |
| 8 | +The 3.1 release of Tarantool continues the development of a new cluster configuration approach introduced in the :ref:`3.0 version <3-0-new_declarative_configuration>` and adds the following main product features and improvements for the Community and Enterprise editions: |
| 9 | + |
| 10 | +* **Community Edition (CE)** |
| 11 | + |
| 12 | + * Improved developer experience for handling errors using the ``box.error`` module. |
| 13 | + * Introduced fixed-size numeric field types: ``uint8``, ``int8``, ``uint16``, and more. |
| 14 | + * Added RPC functionality for accessing custom roles from the configuration. |
| 15 | + * Made the ``tt`` utility used to manage instances fully compatible with the latest Tarantool version. |
| 16 | + |
| 17 | +* **Enterprise Edition (EE)** |
| 18 | + |
| 19 | + * Introduced an external coordinator for automatic and manual failover. |
| 20 | + * Improved the stability of work with the centralized configuration stored in etcd. |
| 21 | + |
| 22 | + |
| 23 | +.. _3-1-features-for-developers: |
| 24 | + |
| 25 | +Developing applications |
| 26 | +----------------------- |
| 27 | + |
| 28 | +.. _3-1-error-handling: |
| 29 | + |
| 30 | +Error handling |
| 31 | +~~~~~~~~~~~~~~ |
| 32 | + |
| 33 | +This release improves the developer experience for handling errors using the :ref:`box.error <box-error-submodule>` module. |
| 34 | +Below are listed the most notable features and changes. |
| 35 | + |
| 36 | + |
| 37 | +.. _3-1-error_payload_fields: |
| 38 | + |
| 39 | +Error payload fields |
| 40 | +******************** |
| 41 | + |
| 42 | +With the 3.1 release, you can add a custom payload to an error. |
| 43 | +The payload is passed as key-value pairs where a key is a string and a value is any Lua object. |
| 44 | +In the example below, the ``description`` key is used to keep the custom payload. |
| 45 | + |
| 46 | +.. code-block:: lua |
| 47 | +
|
| 48 | + custom_error = box.error.new({ type = 'CustomInternalError', |
| 49 | + message = 'Internal server error', |
| 50 | + description = 'Some error details' -- payload |
| 51 | + }) |
| 52 | +
|
| 53 | +A payload field value can be accessed using the dot syntax: |
| 54 | + |
| 55 | +.. code-block:: tarantoolsession |
| 56 | +
|
| 57 | + tarantool> custom_error.description |
| 58 | + --- |
| 59 | + - Some error details |
| 60 | + ... |
| 61 | +
|
| 62 | +
|
| 63 | +
|
| 64 | +
|
| 65 | +
|
| 66 | +.. _3-1-error_stack: |
| 67 | + |
| 68 | +Error stacks |
| 69 | +************ |
| 70 | + |
| 71 | +The 3.1 release simplifies creating error chains. |
| 72 | +In the earlier versions, you need to set an error cause using the :ref:`set_prev(error_object) <box_error-set_prev>` method, for example: |
| 73 | + |
| 74 | +.. code-block:: lua |
| 75 | +
|
| 76 | + local ok, err = pcall(my_func) |
| 77 | + if not ok then |
| 78 | + local err2 = box.error.new{type = "MyAppError", message = "my_func failed"} |
| 79 | + err2:set_prev(err) |
| 80 | + err2:raise() |
| 81 | + end |
| 82 | +
|
| 83 | +Using this approach, you need to construct a new error without raising it, then set its cause using ``set_prev()``, and only then raise it. |
| 84 | +Starting with the 3.1 version, you can use a new ``prev`` argument when constructing an error: |
| 85 | + |
| 86 | +.. code-block:: lua |
| 87 | +
|
| 88 | + local ok, err = pcall(my_func) |
| 89 | + if not ok then |
| 90 | + box.error{type = "MyAppError", message = "my_func failed", prev = err} |
| 91 | + end |
| 92 | +
|
| 93 | +
|
| 94 | +.. _3-1-error_serialization: |
| 95 | + |
| 96 | +Error serialization improvements |
| 97 | +******************************** |
| 98 | + |
| 99 | +The 3.1 release allows you to increase the verbosity of error serialization. |
| 100 | +Before the 3.1 release, a serialized error representation included only an error message: |
| 101 | + |
| 102 | +.. code-block:: tarantoolsession |
| 103 | +
|
| 104 | + tarantool> box.error.new({ type = 'CustomInternalError', message = 'Internal server error'}) |
| 105 | + --- |
| 106 | + - Internal server error |
| 107 | + ... |
| 108 | +
|
| 109 | +
|
| 110 | +Starting with the 3.1 version, a serialized error also includes other fields that might be useful for analyzing errors: |
| 111 | + |
| 112 | +.. code-block:: lua |
| 113 | +
|
| 114 | + tarantool> box.error.new({ type = 'CustomInternalError', message = 'Internal server error'}) |
| 115 | + --- |
| 116 | + - code: 0 |
| 117 | + base_type: CustomError |
| 118 | + type: CustomInternalError |
| 119 | + custom_type: CustomInternalError |
| 120 | + message: Internal server error |
| 121 | + trace: |
| 122 | + - file: '[C]' |
| 123 | + line: 4294967295 |
| 124 | + ... |
| 125 | +
|
| 126 | +Logging an error using a built-in :ref:`logging module <log-module>` prints an error message followed by a tab space (``\t``) and all the payload fields serialized as a JSON map, for example: |
| 127 | + |
| 128 | +.. code-block:: none |
| 129 | +
|
| 130 | + main/104/app.lua/tarantool I> Internal server error {"code":0,"base_type":"CustomError","type":"CustomInternalError", ... } |
| 131 | +
|
| 132 | +Given that this change may change the behavior of existing code, a new ``box_error_serialize_verbose`` :ref:`compat <compat-module>` option is introduced. |
| 133 | +To try out an increased verbosity of error serialization, set this option to ``new``: |
| 134 | + |
| 135 | +.. code-block:: tarantoolsession |
| 136 | +
|
| 137 | + tarantool> require('compat').box_error_serialize_verbose = 'new' |
| 138 | + --- |
| 139 | + ... |
| 140 | +
|
| 141 | +
|
| 142 | +.. _3-1-fixed_size_numeric_types: |
| 143 | + |
| 144 | +Fixed-size numeric types |
| 145 | +~~~~~~~~~~~~~~~~~~~~~~~~ |
| 146 | + |
| 147 | +The 3.1 release introduces fixed-size numeric :ref:`types <index-box_data-types>` that might be useful to store data unencoded in an array for effective scanning. |
| 148 | +The following numeric types are added: |
| 149 | + |
| 150 | +* ``uint8``: an integer in a range ``[0 .. 255]``. |
| 151 | +* ``int8``: an integer in a range ``[-128 .. 127]``. |
| 152 | +* ``uint16``: an integer in a range ``[0 .. 65,535]``. |
| 153 | +* ``int16``: an integer in a range ``[-32,768 .. 32,767]``. |
| 154 | +* ``uint32``: an integer in a range ``[0 .. 4,294,967,295]``. |
| 155 | +* ``int32``: an integer in a range ``[-2,147,483,648 .. 2,147,483,647]``. |
| 156 | +* ``uint64``: an integer in a range ``[0 .. 18,446,744,073,709,551,615]``. |
| 157 | +* ``int64``: an integer in a range ``[-9,223,372,036,854,775,808 .. 9,223,372,036,854,775,807]``. |
| 158 | +* ``float32``: a 32-bit floating point number. |
| 159 | +* ``float64``: a 64-bit floating point number. |
| 160 | + |
| 161 | + |
| 162 | + |
| 163 | +.. _3-1-accessing_configuration: |
| 164 | + |
| 165 | +Accessing configuration of other cluster members |
| 166 | +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 167 | + |
| 168 | +In Tarantool 3.0, the :ref:`config <config-module>` module provides the ability to work with a current instance's configuration only. |
| 169 | +Starting with the 3.1 version, you can get all the instances that constitute a cluster and obtain the configuration of any instance of this cluster. |
| 170 | + |
| 171 | +The ``config:instances()`` function lists all instances of the cluster: |
| 172 | + |
| 173 | +.. code-block:: tarantoolsession |
| 174 | +
|
| 175 | + sharded_cluster:router-a-001> require('config'):instances() |
| 176 | + --- |
| 177 | + - storage-a-001: |
| 178 | + group_name: storages |
| 179 | + instance_name: storage-a-001 |
| 180 | + replicaset_name: storage-a |
| 181 | + storage-b-002: |
| 182 | + group_name: storages |
| 183 | + instance_name: storage-b-002 |
| 184 | + replicaset_name: storage-b |
| 185 | + router-a-001: |
| 186 | + group_name: routers |
| 187 | + instance_name: router-a-001 |
| 188 | + replicaset_name: router-a |
| 189 | + storage-a-002: |
| 190 | + group_name: storages |
| 191 | + instance_name: storage-a-002 |
| 192 | + replicaset_name: storage-a |
| 193 | + storage-b-001: |
| 194 | + group_name: storages |
| 195 | + instance_name: storage-b-001 |
| 196 | + replicaset_name: storage-b |
| 197 | + ... |
| 198 | +
|
| 199 | +To get the specified configuration value for a certain instance, pass an instance name as an argument to ``config:get()``: |
| 200 | + |
| 201 | +.. code-block:: tarantoolsession |
| 202 | +
|
| 203 | + sharded_cluster:router-a-001> require('config'):get('iproto', {instance = 'storage-b-001'}) |
| 204 | + --- |
| 205 | + - readahead: 16320 |
| 206 | + net_msg_max: 768 |
| 207 | + listen: |
| 208 | + - uri: 127.0.0.1:3304 |
| 209 | + threads: 1 |
| 210 | + advertise: |
| 211 | + peer: |
| 212 | + login: replicator |
| 213 | + client: null |
| 214 | + sharding: |
| 215 | + login: storage |
| 216 | + ... |
| 217 | +
|
| 218 | +
|
| 219 | +
|
| 220 | +.. _3-1-administration-and-maintenance: |
| 221 | + |
| 222 | +Administration and maintenance |
| 223 | +------------------------------ |
| 224 | + |
| 225 | +.. _3-1-failover_coordinator: |
| 226 | + |
| 227 | +Failover coordinator (EE) |
| 228 | +~~~~~~~~~~~~~~~~~~~~~~~~~ |
| 229 | + |
| 230 | +Tarantool Enterprise Edition 3.1 introduces an external failover coordinator that monitors a Tarantool cluster and performs automatic leadership change if a current replica set leader is inaccessible. |
| 231 | + |
| 232 | +A failover coordinator requires the :ref:`replication.failover <configuration_reference_replication_failover>` configuration option to be set to ``supervised``: |
| 233 | + |
| 234 | +.. code-block:: yaml |
| 235 | +
|
| 236 | + replication: |
| 237 | + failover: supervised |
| 238 | +
|
| 239 | + # ... |
| 240 | +
|
| 241 | +To start a failover coordinator, execute the ``tarantool`` command with the ``failover`` option and pass a path to a :ref:`YAML configuration file <configuration_overview>`: |
| 242 | + |
| 243 | +.. code-block:: console |
| 244 | +
|
| 245 | + $ tarantool --failover --config /path/to/config |
| 246 | +
|
| 247 | +A failover coordinator connects to all the instances, polls them for their status, and controls that each replica set with ``replication.failover`` set to ``supervised`` has only one writable instance. |
| 248 | + |
| 249 | +Optionally, you can configure failover timeouts and other parameters in the ``failover`` section at the :ref:`global level <configuration_scopes>`: |
| 250 | + |
| 251 | +.. code-block:: yaml |
| 252 | +
|
| 253 | + failover: |
| 254 | + call_timeout: 1 |
| 255 | + lease_interval: 15 |
| 256 | + renew_interval: 5 |
| 257 | + stateboard: |
| 258 | + renew_interval: 1 |
| 259 | + keepalive_interval: 5 |
| 260 | +
|
| 261 | +
|
| 262 | +.. _3-1-sharding: |
| 263 | + |
| 264 | +Sharding |
| 265 | +~~~~~~~~ |
| 266 | + |
| 267 | +The 3.1 release includes new :ref:`sharding <configuration_reference_sharding>` options that provide additional flexibility for configuring a sharded cluster. |
| 268 | +A new ``sharding.weight`` specifies the relative amount of data that a replica set can store. |
| 269 | +In the example below, the ``storage-a`` replica set can store twice as much data as ``storage-b``: |
| 270 | + |
| 271 | +.. code-block:: yaml |
| 272 | +
|
| 273 | + # ... |
| 274 | + replicasets: |
| 275 | + storage-a: |
| 276 | + sharding: |
| 277 | + weight: 2 |
| 278 | + # ... |
| 279 | + storage-b: |
| 280 | + sharding: |
| 281 | + weight: 1 |
| 282 | + # ... |
| 283 | +
|
| 284 | +
|
| 285 | +
|
| 286 | +The ``sharding.rebalancer_mode`` option configures whether a rebalancer is selected manually or automatically. |
| 287 | +This option can have one of three values: |
| 288 | + |
| 289 | +* ``auto`` (default): if there are no replica sets with the ``rebalancer`` sharding role (:ref:`sharding.roles <configuration_reference_sharding_roles>`), a replica set with the rebalancer will be selected automatically among all replica sets. |
| 290 | +* ``manual``: one of the replica sets should have the ``rebalancer`` sharding role. The rebalancer will be in this replica set. |
| 291 | +* ``off``: rebalancing is turned off regardless of whether a replica set with the ``rebalancer`` sharding role exists or not. |
| 292 | + |
| 293 | + |
| 294 | +.. _3-1-compatibility_tt: |
| 295 | + |
| 296 | +Compatibility with the tt utility |
| 297 | +--------------------------------- |
| 298 | + |
| 299 | +With this release, the ``tarantoolctl`` utility used to administer Tarantool instances is completely removed from Tarantool packages. |
| 300 | +The latest version of the :ref:`tt utility <tt-cli>` is fully compatible with Tarantool 3.1 and covers all the required functionality: |
| 301 | + |
| 302 | +* Setting up a development environment: initializing the environment and installing different Tarantool versions. |
| 303 | +* Various capabilities for developing cluster applications: creating applications from templates, managing modules, and building and packaging applications. |
| 304 | +* Managing cluster instances: starting and stopping instances, connecting to remote instances for administration, and so on. |
| 305 | +* Importing and exporting data (Enterprise Edition only). |
| 306 | + |
| 307 | +Learn how to migrate from ``tarantoolctl`` to ``tt`` in the :ref:`tarantoolctl-migration-to-tt` section. |
0 commit comments