|
| 1 | +# IoTDB |
| 2 | + |
| 3 | +> IoTDB数据接收器 |
| 4 | +
|
| 5 | +## 支持引擎 |
| 6 | + |
| 7 | +> Spark<br/> |
| 8 | +> Flink<br/> |
| 9 | +> SeaTunnel Zeta<br/> |
| 10 | +
|
| 11 | +## 描述 |
| 12 | + |
| 13 | +将数据写入IoTDB。 |
| 14 | + |
| 15 | +## Using Dependency |
| 16 | + |
| 17 | +### 适用于Spark/Flink引擎 |
| 18 | + |
| 19 | +> 1. 您需要确保[jdbc驱动程序jar包](https://mvnrepository.com/artifact/org.apache.iotdb/iotdb-jdbc)已放置在目录`${SEATUNNEL_HOME}/plugins/`中。 |
| 20 | +
|
| 21 | +### 适用于SeaTunnelZeta引擎 |
| 22 | + |
| 23 | +> 1. 您需要确保[jdbc驱动程序jar包](https://mvnrepository.com/artifact/org.apache.iotdb/iotdb-jdbc)已放置在目录“${SEATUNNEL_HOME}/lib/”中。 |
| 24 | +
|
| 25 | +## 主要特性 |
| 26 | + |
| 27 | +- [x] [exactly-once](../../concept/connector-v2-features.md) |
| 28 | + |
| 29 | +IoTDB通过幂等写支持 `exactly-once` 功能。如果两条数据 |
| 30 | +如果使用相同的`key`和`timestamp`,新数据将覆盖旧数据。 |
| 31 | + |
| 32 | +:::提示 |
| 33 | + |
| 34 | +IoTDB和Spark之间存在节俭版本冲突。因此,您需要执行`rm -f $SPARK_HOME/jars/libthrift*` 和 `cp $IOTDB_HOME/lib/libthrift* $SPARK_HOME/jars/`来解决这个问题。 |
| 35 | + |
| 36 | +::: |
| 37 | + |
| 38 | +## 支持的数据源信息 |
| 39 | + |
| 40 | +| 数据源 | Supported 版本 | 地址 | |
| 41 | +|------------|--------------------|----------------| |
| 42 | +| IoTDB | `>= 0.13.0` | localhost:6667 | |
| 43 | + |
| 44 | +## 数据类型映射 |
| 45 | + |
| 46 | +| IotDB 数据类型 | SeaTunnel 数据类型 | |
| 47 | +|-----------------|---------------------| |
| 48 | +| BOOLEAN | BOOLEAN | |
| 49 | +| INT32 | TINYINT | |
| 50 | +| INT32 | SMALLINT | |
| 51 | +| INT32 | INT | |
| 52 | +| INT64 | BIGINT | |
| 53 | +| FLOAT | FLOAT | |
| 54 | +| DOUBLE | DOUBLE | |
| 55 | +| TEXT | STRING | |
| 56 | + |
| 57 | +## Sink 选项 |
| 58 | + |
| 59 | +| 名称 | 类型 | 是否必传 | 默认值 | 描述 | |
| 60 | +|-----------------------------|---------|----------|--------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------| |
| 61 | +| node_urls | String | 是 | - | `IoTDB` 集群地址,格式为 `"host1:port"` 或 `"host1:port,host2:port"` | |
| 62 | +| username | String | 是 | - | `IoTDB` 用户的用户名 | |
| 63 | +| password | String | 是 | - | `IoTDB` 用户的密码 | |
| 64 | +| key_device | String | 是 | - | 在SeaTunnelRow中指定`IoTDB`设备ID的字段名 | |
| 65 | +| key_timestamp | String | 否 | processing time | 在SeaTunnelRow中指定`IoTDB`时间戳的字段名。如果未指定,则使用处理时间作为时间戳 | |
| 66 | +| key_measurement_fields | Array | 否 | exclude `device` & `timestamp` | 在SeaTunnelRow中指定`IoTDB`测量列表的字段名称。如果未指定,则包括所有字段,但排除 `device` & `timestamp` | |
| 67 | +| storage_group | Array | 否 | - | 指定设备存储组(路径前缀) <br/> 例如: deviceId = ${storage_group} + "." + ${key_device} | |
| 68 | +| batch_size | Integer | 否 | 1024 | 对于批写入,当缓冲区的数量达到`batch_size`的数量或时间达到`batch_interval_ms`时,数据将被刷新到IoTDB中 | |
| 69 | +| max_retries | Integer | 否 | - | 刷新的重试次数 failed | |
| 70 | +| retry_backoff_multiplier_ms | Integer | 否 | - | 用作生成下一个退避延迟的乘数 | |
| 71 | +| max_retry_backoff_ms | Integer | 否 | - | 尝试重试对`IoTDB`的请求之前等待的时间量 | |
| 72 | +| default_thrift_buffer_size | Integer | 否 | - | 在`IoTDB`客户端中节省初始化缓冲区大小 | |
| 73 | +| max_thrift_frame_size | Integer | 否 | - | 在`IoTDB`客户端中节约最大帧大小 | |
| 74 | +| zone_id | string | 否 | - | `IoTDB` java.time.ZoneId client | |
| 75 | +| enable_rpc_compression | Boolean | 否 | - | 在`IoTDB`客户端中启用rpc压缩 | |
| 76 | +| connection_timeout_in_ms | Integer | 否 | - | 连接到`IoTDB`时等待的最长时间(毫秒) | |
| 77 | +| common-options | | 否 | - | Sink插件常用参数,详见[Sink common Options](../Sink common Options.md) | |
| 78 | + |
| 79 | +## 示例 |
| 80 | + |
| 81 | +```hocon |
| 82 | +env { |
| 83 | + parallelism = 2 |
| 84 | + job.mode = "BATCH" |
| 85 | +} |
| 86 | +
|
| 87 | +source { |
| 88 | + FakeSource { |
| 89 | + row.num = 16 |
| 90 | + bigint.template = [1664035200001] |
| 91 | + schema = { |
| 92 | + fields { |
| 93 | + device_name = "string" |
| 94 | + temperature = "float" |
| 95 | + moisture = "int" |
| 96 | + event_ts = "bigint" |
| 97 | + c_string = "string" |
| 98 | + c_boolean = "boolean" |
| 99 | + c_tinyint = "tinyint" |
| 100 | + c_smallint = "smallint" |
| 101 | + c_int = "int" |
| 102 | + c_bigint = "bigint" |
| 103 | + c_float = "float" |
| 104 | + c_double = "double" |
| 105 | + } |
| 106 | + } |
| 107 | + } |
| 108 | +} |
| 109 | +``` |
| 110 | + |
| 111 | +上游SeaTunnelRow数据格式如下: |
| 112 | + |
| 113 | +| device_name | temperature | moisture | event_ts | c_string | c_boolean | c_tinyint | c_smallint | c_int | c_bigint | c_float | c_double | |
| 114 | +|--------------------------|-------------|----------|---------------|----------|-----------|-----------|------------|-------|------------|---------|----------| |
| 115 | +| root.test_group.device_a | 36.1 | 100 | 1664035200001 | abc1 | true | 1 | 1 | 1 | 2147483648 | 1.0 | 1.0 | |
| 116 | +| root.test_group.device_b | 36.2 | 101 | 1664035200001 | abc2 | false | 2 | 2 | 2 | 2147483649 | 2.0 | 2.0 | |
| 117 | +| root.test_group.device_c | 36.3 | 102 | 1664035200001 | abc3 | false | 3 | 3 | 3 | 2147483649 | 3.0 | 3.0 | |
| 118 | + |
| 119 | +### 案例1 |
| 120 | + |
| 121 | +只填写所需的配置。 |
| 122 | +使用当前处理时间作为时间戳。并包括所有字段,但不包括`device` & `timestamp`作为测量字段 |
| 123 | + |
| 124 | +```hocon |
| 125 | +sink { |
| 126 | + IoTDB { |
| 127 | + node_urls = "localhost:6667" |
| 128 | + username = "root" |
| 129 | + password = "root" |
| 130 | + key_device = "device_name" # specify the `deviceId` use device_name field |
| 131 | + } |
| 132 | +} |
| 133 | +``` |
| 134 | + |
| 135 | +"IoTDB"数据格式的输出如下: |
| 136 | + |
| 137 | +```shell |
| 138 | +IoTDB> SELECT * FROM root.test_group.* align by device; |
| 139 | ++------------------------+------------------------+--------------+-----------+--------------+---------+----------+----------+-----------+------+-----------+--------+---------+ |
| 140 | +| Time| Device| temperature| moisture| event_ts| c_string| c_boolean| c_tinyint| c_smallint| c_int| c_bigint| c_float| c_double| |
| 141 | ++------------------------+------------------------+--------------+-----------+--------------+---------+----------+----------+-----------+------+-----------+--------+---------+ |
| 142 | +|2023-09-01T00:00:00.001Z|root.test_group.device_a| 36.1| 100| 1664035200001| abc1| true| 1| 1| 1| 2147483648| 1.0| 1.0| |
| 143 | +|2023-09-01T00:00:00.001Z|root.test_group.device_b| 36.2| 101| 1664035200001| abc2| false| 2| 2| 2| 2147483649| 2.0| 2.0| |
| 144 | +|2023-09-01T00:00:00.001Z|root.test_group.device_c| 36.3| 102| 1664035200001| abc2| false| 3| 3| 3| 2147483649| 3.0| 3.0| |
| 145 | ++------------------------+------------------------+--------------+-----------+--------------+---------+---------+-----------+-----------+------+-----------+--------+---------+ |
| 146 | +``` |
| 147 | + |
| 148 | +### 案例2 |
| 149 | + |
| 150 | +使用源事件的时间 |
| 151 | + |
| 152 | +```hocon |
| 153 | +sink { |
| 154 | + IoTDB { |
| 155 | + node_urls = "localhost:6667" |
| 156 | + username = "root" |
| 157 | + password = "root" |
| 158 | + key_device = "device_name" # specify the `deviceId` use device_name field |
| 159 | + key_timestamp = "event_ts" # specify the `timestamp` use event_ts field |
| 160 | + } |
| 161 | +} |
| 162 | +``` |
| 163 | + |
| 164 | +"IoTDB"数据格式的输出如下: |
| 165 | + |
| 166 | +```shell |
| 167 | +IoTDB> SELECT * FROM root.test_group.* align by device; |
| 168 | ++------------------------+------------------------+--------------+-----------+--------------+---------+----------+----------+-----------+------+-----------+--------+---------+ |
| 169 | +| Time| Device| temperature| moisture| event_ts| c_string| c_boolean| c_tinyint| c_smallint| c_int| c_bigint| c_float| c_double| |
| 170 | ++------------------------+------------------------+--------------+-----------+--------------+---------+----------+----------+-----------+------+-----------+--------+---------+ |
| 171 | +|2022-09-25T00:00:00.001Z|root.test_group.device_a| 36.1| 100| 1664035200001| abc1| true| 1| 1| 1| 2147483648| 1.0| 1.0| |
| 172 | +|2022-09-25T00:00:00.001Z|root.test_group.device_b| 36.2| 101| 1664035200001| abc2| false| 2| 2| 2| 2147483649| 2.0| 2.0| |
| 173 | +|2022-09-25T00:00:00.001Z|root.test_group.device_c| 36.3| 102| 1664035200001| abc2| false| 3| 3| 3| 2147483649| 3.0| 3.0| |
| 174 | ++------------------------+------------------------+--------------+-----------+--------------+---------+---------+-----------+-----------+------+-----------+--------+---------+ |
| 175 | +``` |
| 176 | + |
| 177 | +### 案例3 |
| 178 | + |
| 179 | +使用源事件的时间和限制度量字段 |
| 180 | + |
| 181 | +```hocon |
| 182 | +sink { |
| 183 | + IoTDB { |
| 184 | + node_urls = "localhost:6667" |
| 185 | + username = "root" |
| 186 | + password = "root" |
| 187 | + key_device = "device_name" |
| 188 | + key_timestamp = "event_ts" |
| 189 | + key_measurement_fields = ["temperature", "moisture"] |
| 190 | + } |
| 191 | +} |
| 192 | +``` |
| 193 | + |
| 194 | +"IoTDB"数据格式的输出如下: |
| 195 | + |
| 196 | +```shell |
| 197 | +IoTDB> SELECT * FROM root.test_group.* align by device; |
| 198 | ++------------------------+------------------------+--------------+-----------+ |
| 199 | +| Time| Device| temperature| moisture| |
| 200 | ++------------------------+------------------------+--------------+-----------+ |
| 201 | +|2022-09-25T00:00:00.001Z|root.test_group.device_a| 36.1| 100| |
| 202 | +|2022-09-25T00:00:00.001Z|root.test_group.device_b| 36.2| 101| |
| 203 | +|2022-09-25T00:00:00.001Z|root.test_group.device_c| 36.3| 102| |
| 204 | ++------------------------+------------------------+--------------+-----------+ |
| 205 | +``` |
| 206 | + |
| 207 | +## 更改日志 |
| 208 | + |
| 209 | +### 2.2.0-beta 2022-09-26 |
| 210 | + |
| 211 | +- 添加IoTDB数据接收器 |
| 212 | + |
| 213 | +### 2.3.0-beta 2022-10-20 |
| 214 | + |
| 215 | +- [Improve] 改进IoTDB数据接收器 ([2917](https://github.com/apache/seatunnel/pull/2917)) |
| 216 | + - 支持sql语法对齐 |
| 217 | + - 支持sql拆分忽略案例 |
| 218 | + - 支持将拆分偏移量恢复到至少一次 |
| 219 | + - 支持从RowRecord读取时间戳 |
| 220 | +- [BugFix] 固定IoTDB连接器写入NPE ([3080](https://github.com/apache/seatunnel/pull/3080)) |
| 221 | + |
0 commit comments