Installing ThingsBoard PE on Windows

ThingsBoard团队建议将PostgreSQL用于负载（<5000消息/秒）的开发和生产环境，使用公有云托管的PostgreSQL数据库服务对于某些ThingsBoard实例而言是一种经济高效的方式。

PostgreSQL安装

下载安装文件(PostgreSQL 11.7或更高版本)此处并按照安装说明进行操作。

在PostgreSQL安装时系统将提示您输入超级用户(postgres)密码。

请一定牢记此密码，为了方便记忆我们将密码替换为postgres。

创建ThingsBoard数据库

安装成功后启动”pgAdmin”并使用超级用户(postgres)身份登录。
打开服务器并用”postgres”用户创建数据库”thingsboard”。

ThingsBoard配置

如果已将PostgreSQL超级用户密码指定为”postgres”则可以跳过此步骤。

以管理员用户身份打开记事本或其他编辑器（右键单击应用程序图标然后选择”以管理员身份运行”）。
打开以下文件进行编辑（在文件选择对话框中选择“所有文件”而不是“文本文档”，编码为UTF-8）：

1
C:\Program Files (x86)\thingsboard\conf\thingsboard.yml

找到”SQL DAO Configuration”代码块将postgres用户密码替换”postgres”：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
# SQL DAO Configuration
spring:
  data:
    jpa:
      repositories:
        enabled: "true"
  jpa:
    open-in-view: "false"
    hibernate:
      ddl-auto: "none"
  datasource:
    driverClassName: "${SPRING_DRIVER_CLASS_NAME:org.postgresql.Driver}"
    url: "${SPRING_DATASOURCE_URL:jdbc:postgresql://localhost:5432/thingsboard}"
    username: "${SPRING_DATASOURCE_USERNAME:postgres}"
    password: "${SPRING_DATASOURCE_PASSWORD:YOUR_POSTGRES_PASSWORD_HERE}"
    hikari:
      maximumPoolSize: "${SPRING_DATASOURCE_MAXIMUM_POOL_SIZE:5}"

找到”SQL_POSTGRES_TS_KV_PARTITIONING”参数以覆盖时间戳键值存储分区大小的默认值：

1
2
3
4
    sql:
      postgres:
        # Specify partitioning size for timestamp key-value storage. Example: DAYS, MONTHS, YEARS, INDEFINITE.
        ts_key_value_partitioning: "${SQL_POSTGRES_TS_KV_PARTITIONING:MONTHS}"

ThingsBoard团队建议具有在生产环境使用TimescaleDB经验的公司使用Timescale数据库把时间序列数据存储在TimescaleDB Hypertable中并将PostgreSQL用于主要实体（设备/资产/仪表板/客户）。

PostgreSQL安装

下载安装文件(PostgreSQL 11.7或更高版本)此处并按照安装说明进行操作。

在PostgreSQL安装时系统将提示您输入超级用户(postgres)密码。

请一定牢记此密码，为了方便记忆我们将密码替换为postgres。

创建ThingsBoard数据库

安装成功后启动”pgAdmin”并使用超级用户(postgres)身份登录。
打开服务器并用”postgres”用户创建数据库”thingsboard”。

ThingsBoard配置

以管理员用户身份打开记事本或其他编辑器（右键单击应用程序图标然后选择”以管理员身份运行”）。
打开以下文件进行编辑（在文件选择对话框中选择“所有文件”而不是“文本文档”，编码为UTF-8）：

1
C:\Program Files (x86)\thingsboard\conf\thingsboard.yml

找到”SQL DAO Configuration”代码块将postgres用户密码替换”postgres”：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
# SQL DAO Configuration
spring:
  data:
    jpa:
      repositories:
        enabled: "true"
  jpa:
    open-in-view: "false"
    hibernate:
      ddl-auto: "none"
  datasource:
    driverClassName: "${SPRING_DRIVER_CLASS_NAME:org.postgresql.Driver}"
    url: "${SPRING_DATASOURCE_URL:jdbc:postgresql://localhost:5432/thingsboard}"
    username: "${SPRING_DATASOURCE_USERNAME:postgres}"
    password: "${SPRING_DATASOURCE_PASSWORD:YOUR_POSTGRES_PASSWORD_HERE}"
    hikari:
      maximumPoolSize: "${SPRING_DATASOURCE_MAXIMUM_POOL_SIZE:5}"

找到”DATABASE_TS_TYPE”参数将”sql”替换为”timescale”。

1
2
3
4
      ts:
        type: "${DATABASE_TS_TYPE:sql}" # cassandra, sql, or timescale (for hybrid mode, DATABASE_TS_TYPE value should be cassandra, or timescale)
    
    # note: timescale works only with postgreSQL database for DATABASE_ENTITIES_TYPE.

可以选择调整引用Timescale DB配置的参数”sql”配置块中的”timescale”配置块。

1
2
3
4
5
# SQL configuration parameters
sql:
    timescale:
      # Specify Interval size for new data chunks storage.
      chunk_time_interval: "${SQL_TIMESCALE_CHUNK_TIME_INTERVAL:604800000}"

默认使用内存队列无需其他配置。

Kafka安装

Apache Kafka是一个开源的流式数据处理平台。

Kafka安装

使用此说明在Docker容器中安装Kafka。

ThingsBoard配置

以管理员用户身份打开记事本或其他编辑器（右键单击应用程序图标，然后选择“以管理员身份运行”）。
打开以下文件进行编辑（在文件选择对话框中选择“所有文件”而不是“文本文档”，编码为UTF-8）：

1
C:\Program Files (x86)\thingsboard\conf\thingsboard.yml

and locate “queue:” block. Make sure the queue type is “kafka”, and don’t forget to replace “localhost:9092” with your real Kafka bootstrap servers:

1
2
3
4
5
6
7
8
9
10
11
queue:
  type: "${TB_QUEUE_TYPE:kafka}"
...
  kafka:
    bootstrap.servers: "${TB_KAFKA_SERVERS:localhost:9092}"
    acks: "${TB_KAFKA_ACKS:all}"
    retries: "${TB_KAFKA_RETRIES:1}"
    batch.size: "${TB_KAFKA_BATCH_SIZE:16384}"
    linger.ms: "${TB_KAFKA_LINGER_MS:1}"
    buffer.memory: "${TB_BUFFER_MEMORY:33554432}"
    replication_factor: "${TB_QUEUE_KAFKA_REPLICATION_FACTOR:1}"

AWS SQS配置

首先需要创建AWS账户然后访问AWS SQS服务。

使用此说明创建AWS SQS服务凭证。

• Access key ID
• Secret access key

ThingsBoard配置

以管理员用户身份打开记事本或其他编辑器（右键单击应用程序图标，然后选择“以管理员身份运行”）。
打开以下文件进行编辑（在文件选择对话框中选择“所有文件”而不是“文本文档”，编码为UTF-8）：

1
C:\Program Files (x86)\thingsboard\conf\thingsboard.yml

添加配置将”YOUR_KEY”和”YOUR_SECRET”替换为真实的AWS用户凭证并将”YOUR_REGION”替换成AWS SQS帐户区域：and “YOUR_REGION” with your real AWS SQS account region:

1
2
3
4
5
6
7
queue:
  type: "${TB_QUEUE_TYPE:aws-sqs}"
...
  aws_sqs:
    access_key_id: "${TB_QUEUE_AWS_SQS_ACCESS_KEY_ID:YOUR_KEY}"
    secret_access_key: "${TB_QUEUE_AWS_SQS_SECRET_ACCESS_KEY:YOUR_SECRET}"
    region: "${TB_QUEUE_AWS_SQS_REGION:YOUR_REGION}"

These params affect the number of requests per second from each partitions per each queue.

Number of requests to particular Message Queue calculated based on the formula:

((Number of Rule Engine and Core Queues) * (Number of partitions per Queue) + (Number of transport queues) + (Number of microservices) + (Number of JS executors)) * 1000 / POLL_INTERVAL_MS

For example, number of requests based on default parameters is:

Rule Engine queues:

Main 10 partitions + HighPriority 10 partitions + SequentialByOriginator 10 partitions = 30

Core queue 10 partitions

Transport request Queue + response Queue = 2

Rule Engine Transport notifications Queue + Core Transport notifications Queue = 2

Total = 44

Number of requests per second = 44 * 1000 / 25 = 1760 requests

Based on the use case, you can compromise latency and decrease number of partitions/requests to the queue, if the message load is low.

By UI set the parameters - interval (1000) and partitions (1) for Rule Engine queues.

Sample parameters to fit into 10 requests per second on a “monolith” deployment:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
queue:
...
  transport_api:
    request_poll_interval: "${TB_QUEUE_TRANSPORT_REQUEST_POLL_INTERVAL_MS:1000}"
    response_poll_interval: "${TB_QUEUE_TRANSPORT_RESPONSE_POLL_INTERVAL_MS:1000}"
...
  core:
    poll-interval: "${TB_QUEUE_CORE_POLL_INTERVAL_MS:1000}"
    partitions: "${TB_QUEUE_CORE_PARTITIONS:2}"
...    
  vc:
    partitions: "${TB_QUEUE_VC_PARTITIONS:1}"
    poll-interval: "${TB_QUEUE_VC_INTERVAL_MS:1000}"
...
  js:
    response_poll_interval: "${REMOTE_JS_RESPONSE_POLL_INTERVAL_MS:1000}"
...
  rule-engine:
    poll-interval: "${TB_QUEUE_RULE_ENGINE_POLL_INTERVAL_MS:1000}"
...
  transport:
    poll_interval: "${TB_QUEUE_TRANSPORT_NOTIFICATIONS_POLL_INTERVAL_MS:1000}"

可以使用UI更新默认规则引擎队列配置（轮询间隔和分区）有关ThingsBoard规则引擎队列的更多信息请参阅文档。

Google发布/订阅配置

创建一个Google帐户并访问发布/订阅服务。

使用此说明创建一个项目并使用发布/订阅服务。

使用此说明创建服务帐户凭据并编辑角色或管理员后保存json凭据步骤9的文件此处。

ThingsBoard配置

以管理员用户身份打开记事本或其他编辑器（右键单击应用程序图标，然后选择“以管理员身份运行”）。
打开以下文件进行编辑（在文件选择对话框中选择“所有文件”而不是“文本文档”，编码为UTF-8）：

1
C:\Program Files (x86)\thingsboard\conf\thingsboard.yml

and locate “queue:” block. Make sure the queue type is “pubsub”, and don’t forget to replace “YOUR_PROJECT_ID”, “YOUR_SERVICE_ACCOUNT” with your real Pub/Sub project id, and service account (it is whole data from json file):

1
2
3
4
5
6
7
8
queue:
  type: "${TB_QUEUE_TYPE:pubsub}"
...
  pubsub:
    project_id: "${TB_QUEUE_PUBSUB_PROJECT_ID:YOUR_PROJECT_ID}"
    service_account: "${TB_QUEUE_PUBSUB_SERVICE_ACCOUNT:YOUR_SERVICE_ACCOUNT}"
    max_msg_size: "${TB_QUEUE_PUBSUB_MAX_MSG_SIZE:1048576}" #in bytes
    max_messages: "${TB_QUEUE_PUBSUB_MAX_MESSAGES:1000}"

These params affect the number of requests per second from each partitions per each queue.

Number of requests to particular Message Queue calculated based on the formula:

((Number of Rule Engine and Core Queues) * (Number of partitions per Queue) + (Number of transport queues) + (Number of microservices) + (Number of JS executors)) * 1000 / POLL_INTERVAL_MS

For example, number of requests based on default parameters is:

Rule Engine queues:

Main 10 partitions + HighPriority 10 partitions + SequentialByOriginator 10 partitions = 30

Core queue 10 partitions

Transport request Queue + response Queue = 2

Rule Engine Transport notifications Queue + Core Transport notifications Queue = 2

Total = 44

Number of requests per second = 44 * 1000 / 25 = 1760 requests

Based on the use case, you can compromise latency and decrease number of partitions/requests to the queue, if the message load is low.

By UI set the parameters - interval (1000) and partitions (1) for Rule Engine queues.

Sample parameters to fit into 10 requests per second on a “monolith” deployment:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
queue:
...
  transport_api:
    request_poll_interval: "${TB_QUEUE_TRANSPORT_REQUEST_POLL_INTERVAL_MS:1000}"
    response_poll_interval: "${TB_QUEUE_TRANSPORT_RESPONSE_POLL_INTERVAL_MS:1000}"
...
  core:
    poll-interval: "${TB_QUEUE_CORE_POLL_INTERVAL_MS:1000}"
    partitions: "${TB_QUEUE_CORE_PARTITIONS:2}"
...
  vc:
    partitions: "${TB_QUEUE_VC_PARTITIONS:1}"
    poll-interval: "${TB_QUEUE_VC_INTERVAL_MS:1000}"
...
  js:
    response_poll_interval: "${REMOTE_JS_RESPONSE_POLL_INTERVAL_MS:1000}"
...
  rule-engine:
    poll-interval: "${TB_QUEUE_RULE_ENGINE_POLL_INTERVAL_MS:1000}"
...
  transport:
    poll_interval: "${TB_QUEUE_TRANSPORT_NOTIFICATIONS_POLL_INTERVAL_MS:1000}"

可以使用UI更新默认规则引擎队列配置（轮询间隔和分区）有关ThingsBoard规则引擎队列的更多信息请参阅文档。

Azure服务总线配置

创建Azure帐户并访问Azure服务总线。

通过使用说明了解并使用总线服务。

使用说明创建共享访问签名。

ThingsBoard配置

以管理员用户身份打开记事本或其他编辑器（右键单击应用程序图标，然后选择“以管理员身份运行”）。
打开以下文件进行编辑（在文件选择对话框中选择“所有文件”而不是“文本文档”，编码为UTF-8）：

1
C:\Program Files (x86)\thingsboard\conf\thingsboard.yml

添加下面配置内容使用真正的服务总线名称空间替换”YOUR_NAMESPACE_NAME”和”YOUR_SAS_KEY_NAME”及”YOUR_SAS_KEY”：

1
2
3
4
5
6
7
8
queue:
  type: "${TB_QUEUE_TYPE:service-bus}"
...
  service_bus:
    namespace_name: "${TB_QUEUE_SERVICE_BUS_NAMESPACE_NAME:YOUR_NAMESPACE_NAME}"
    sas_key_name: "${TB_QUEUE_SERVICE_BUS_SAS_KEY_NAME:YOUR_SAS_KEY_NAME}"
    sas_key: "${TB_QUEUE_SERVICE_BUS_SAS_KEY:YOUR_SAS_KEY}"
    max_messages: "${TB_QUEUE_SERVICE_BUS_MAX_MESSAGES:1000}"

These params affect the number of requests per second from each partitions per each queue.

Number of requests to particular Message Queue calculated based on the formula:

((Number of Rule Engine and Core Queues) * (Number of partitions per Queue) + (Number of transport queues) + (Number of microservices) + (Number of JS executors)) * 1000 / POLL_INTERVAL_MS

For example, number of requests based on default parameters is:

Rule Engine queues:

Main 10 partitions + HighPriority 10 partitions + SequentialByOriginator 10 partitions = 30

Core queue 10 partitions

Transport request Queue + response Queue = 2

Rule Engine Transport notifications Queue + Core Transport notifications Queue = 2

Total = 44

Number of requests per second = 44 * 1000 / 25 = 1760 requests

Based on the use case, you can compromise latency and decrease number of partitions/requests to the queue, if the message load is low.

By UI set the parameters - interval (1000) and partitions (1) for Rule Engine queues.

Sample parameters to fit into 10 requests per second on a “monolith” deployment:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
queue:
...
  transport_api:
    request_poll_interval: "${TB_QUEUE_TRANSPORT_REQUEST_POLL_INTERVAL_MS:1000}"
    response_poll_interval: "${TB_QUEUE_TRANSPORT_RESPONSE_POLL_INTERVAL_MS:1000}"
...
  core:
    poll-interval: "${TB_QUEUE_CORE_POLL_INTERVAL_MS:1000}"
    partitions: "${TB_QUEUE_CORE_PARTITIONS:2}"
...
  vc:
    partitions: "${TB_QUEUE_VC_PARTITIONS:1}"
    poll-interval: "${TB_QUEUE_VC_INTERVAL_MS:1000}"
...
  js:
    response_poll_interval: "${REMOTE_JS_RESPONSE_POLL_INTERVAL_MS:1000}"
...
  rule-engine:
    poll-interval: "${TB_QUEUE_RULE_ENGINE_POLL_INTERVAL_MS:1000}"
...
  transport:
    poll_interval: "${TB_QUEUE_TRANSPORT_NOTIFICATIONS_POLL_INTERVAL_MS:1000}"

可以使用UI更新默认规则引擎队列配置（轮询间隔和分区）有关ThingsBoard规则引擎队列的更多信息请参阅文档。

RabbitMQ安装

你可以使用官方文档安装RabbitMQ或按照以下说明：

ThingsBoard配置

以管理员用户身份打开记事本或其他编辑器（右键单击应用程序图标，然后选择“以管理员身份运行”）。
打开以下文件进行编辑（在文件选择对话框中选择“所有文件”而不是“文本文档”，编码为UTF-8）：

1
C:\Program Files (x86)\thingsboard\conf\thingsboard.yml

将以下行添加到配置文件将“YOUR_USERNAME”和“YOUR_PASSWORD”替换为真实的信息将“localhost”和“5672”替换为真实的RabbitMQ主机和端口：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
queue:
  type: "${TB_QUEUE_TYPE:rabbitmq}"
...
  rabbitmq:
    exchange_name: "${TB_QUEUE_RABBIT_MQ_EXCHANGE_NAME:}"
    host: "${TB_QUEUE_RABBIT_MQ_HOST:localhost}"
    port: "${TB_QUEUE_RABBIT_MQ_PORT:5672}"
    virtual_host: "${TB_QUEUE_RABBIT_MQ_VIRTUAL_HOST:/}"
    username: "${TB_QUEUE_RABBIT_MQ_USERNAME:YOUR_USERNAME}"
    password: "${TB_QUEUE_RABBIT_MQ_PASSWORD:YOUR_PASSWORD}"
    automatic_recovery_enabled: "${TB_QUEUE_RABBIT_MQ_AUTOMATIC_RECOVERY_ENABLED:false}"
    connection_timeout: "${TB_QUEUE_RABBIT_MQ_CONNECTION_TIMEOUT:60000}"
    handshake_timeout: "${TB_QUEUE_RABBIT_MQ_HANDSHAKE_TIMEOUT:10000}"
    queue-properties:

Confluent配置

你应该创建一个帐户后访问Confluent云然后创建一个Kafka集群和 API Key。

ThingsBoard配置

以管理员用户身份打开记事本或其他编辑器（右键单击应用程序图标，然后选择“以管理员身份运行”）。
打开以下文件进行编辑（在文件选择对话框中选择“所有文件”而不是“文本文档”，编码为UTF-8）：

1
C:\Program Files (x86)\thingsboard\conf\thingsboard.yml

and locate “queue:” block. Make sure the queue type is “kafka”, replication factor is “3” and use confluent cloud is “true”.

Don’t forget to replace “CLUSTER_API_KEY”, “CLUSTER_API_SECRET” and “localhost:9092” with your real Confluent Cloud bootstrap servers:

1
2
3
4
5
6
7
8
9
10
11
queue:
  type: "${TB_QUEUE_TYPE:kafka}"
...
  kafka:
    bootstrap.servers: "${TB_KAFKA_SERVERS:localhost:9092}"
...
    replication_factor: "${TB_QUEUE_KAFKA_REPLICATION_FACTOR:3}"
...
    use_confluent_cloud: "${TB_QUEUE_KAFKA_USE_CONFLUENT_CLOUD:true}"
...
      sasl.config: "${TB_QUEUE_KAFKA_CONFLUENT_SASL_JAAS_CONFIG:org.apache.kafka.common.security.plain.PlainLoginModule required username=\"CLUSTER_API_KEY\" password=\"CLUSTER_API_SECRET\";}"

These params affect the number of requests per second from each partitions per each queue.

Number of requests to particular Message Queue calculated based on the formula:

((Number of Rule Engine and Core Queues) * (Number of partitions per Queue) + (Number of transport queues) + (Number of microservices) + (Number of JS executors)) * 1000 / POLL_INTERVAL_MS

For example, number of requests based on default parameters is:

Rule Engine queues:

Main 10 partitions + HighPriority 10 partitions + SequentialByOriginator 10 partitions = 30

Core queue 10 partitions

Transport request Queue + response Queue = 2

Rule Engine Transport notifications Queue + Core Transport notifications Queue = 2

Total = 44

Number of requests per second = 44 * 1000 / 25 = 1760 requests

Based on the use case, you can compromise latency and decrease number of partitions/requests to the queue, if the message load is low.

By UI set the parameters - interval (1000) and partitions (1) for Rule Engine queues.

Sample parameters to fit into 10 requests per second on a “monolith” deployment:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
queue:
...
  transport_api:
    request_poll_interval: "${TB_QUEUE_TRANSPORT_REQUEST_POLL_INTERVAL_MS:1000}"
    response_poll_interval: "${TB_QUEUE_TRANSPORT_RESPONSE_POLL_INTERVAL_MS:1000}"
...
  core:
    poll-interval: "${TB_QUEUE_CORE_POLL_INTERVAL_MS:1000}"
    partitions: "${TB_QUEUE_CORE_PARTITIONS:2}"
...
  vc:
    partitions: "${TB_QUEUE_VC_PARTITIONS:1}"
    poll-interval: "${TB_QUEUE_VC_INTERVAL_MS:1000}"
...
  js:
    response_poll_interval: "${REMOTE_JS_RESPONSE_POLL_INTERVAL_MS:1000}"
...
  rule-engine:
    poll-interval: "${TB_QUEUE_RULE_ENGINE_POLL_INTERVAL_MS:1000}"
...
  transport:
    poll_interval: "${TB_QUEUE_TRANSPORT_NOTIFICATIONS_POLL_INTERVAL_MS:1000}"

可以使用UI更新默认规则引擎队列配置（轮询间隔和分区）有关ThingsBoard规则引擎队列的更多信息请参阅文档。