turbo-pipelines

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Turbo Pipeline Configuration Reference

Turbo管道配置参考

YAML configuration reference for Turbo pipelines. This is a lookup reference — for interactive pipeline building, use

/turbo-builder

. For pipeline troubleshooting, use

/turbo-doctor

CRITICAL: Always validate YAML with
goldsky turbo validate <file.yaml>
before showing complete pipeline YAML to the user or deploying.

这是Turbo管道的YAML配置参考文档，用于查询查阅——如需交互式管道构建，请使用

/turbo-builder

；如需管道故障排查，请使用

/turbo-doctor

。

重要提示： 在向用户展示完整管道YAML或部署之前，请务必使用
goldsky turbo validate <file.yaml>
验证YAML文件。

Quick Start

快速开始

Deploy a minimal pipeline:

yaml

name: my-first-pipeline
resource_size: s
sources:
  transfers:
    type: dataset
    dataset_name: base.erc20_transfers
    version: 1.2.0
    start_at: latest
transforms: {}
sinks:
  output:
    type: blackhole
    from: transfers

bash

undefined

部署最小化管道：

yaml

name: my-first-pipeline
resource_size: s
sources:
  transfers:
    type: dataset
    dataset_name: base.erc20_transfers
    version: 1.2.0
    start_at: latest
transforms: {}
sinks:
  output:
    type: blackhole
    from: transfers

bash

undefined

Validate first:

先验证：

goldsky turbo validate pipeline.yaml

Then deploy:

然后部署：

goldsky turbo apply pipeline.yaml -i

---

goldsky turbo apply pipeline.yaml -i

---

Prerequisites

前置条件

Goldsky CLI installed -
```
curl https://goldsky.com | sh
```
Turbo CLI extension installed (SEPARATE binary!) -
```
curl https://install-turbo.goldsky.com | sh
```
- Note: Run
```
goldsky turbo list
```
  - if you see "The turbo binary is not installed", install it first
Logged in -
```
goldsky login
```
Pipeline YAML file ready
Secrets created for sinks (if using PostgreSQL, ClickHouse, Kafka, etc.)

已安装Goldsky CLI -
```
curl https://goldsky.com | sh
```
已安装Turbo CLI扩展（独立二进制文件！）-
```
curl https://install-turbo.goldsky.com | sh
```
- 注意：运行
```
goldsky turbo list
```
  ——如果显示"The turbo binary is not installed"，请先安装它
已登录 -
```
goldsky login
```
准备好管道YAML文件
为接收器创建了密钥（如果使用PostgreSQL、ClickHouse、Kafka等）

Discovering Available Data Sources

发现可用数据源

For dataset discovery, invoke the
datasets
skill.

Quick reference for common datasets:

What They Want	Dataset to Use
Token transfers (fungible)	`<chain>.erc20_transfers`
NFT transfers	`<chain>.erc721_transfers`
All contract events	`<chain>.logs`
Block data	`<chain>.blocks`
Transaction data	`<chain>.transactions`

For full chain prefixes, dataset types, and version discovery, use

/datasets

如需发现可用数据集，请调用
datasets
技能。

常见数据集速查：

需求	使用的数据集
同质化代币转账	`<chain>.erc20_transfers`
NFT转账	`<chain>.erc721_transfers`
所有合约事件	`<chain>.logs`
区块数据	`<chain>.blocks`
交易数据	`<chain>.transactions`

如需完整的链前缀、数据集类型及版本信息，请使用

/datasets

。

Quick Reference

速查参考

Installation Commands

安装命令

Action	Command
Install Goldsky CLI	`curl https://goldsky.com \| sh`
Install Turbo extension	`curl https://install-turbo.goldsky.com \| sh`
Verify Turbo installed	`goldsky turbo list`

操作	命令
安装Goldsky CLI	`curl https://goldsky.com \| sh`
安装Turbo扩展	`curl https://install-turbo.goldsky.com \| sh`
验证Turbo是否安装	`goldsky turbo list`

Pipeline Commands

管道命令

Action	Command
List datasets	`goldsky dataset list` ⚠️ Slow (30-60s)
Validate (REQUIRED)	`goldsky turbo validate pipeline.yaml` ✓ Fast (3s)
Deploy/Update	`goldsky turbo apply pipeline.yaml`
Deploy + Inspect	`goldsky turbo apply pipeline.yaml -i`
List pipelines	`goldsky turbo list`
View live data	`goldsky turbo inspect <name>`
Inspect node	`goldsky turbo inspect <name> -n <node>`
View logs	`goldsky turbo logs <name>`
Follow logs	`goldsky turbo logs <name> --follow`
List secrets	`goldsky secret list`

For pause, resume, restart, and delete commands, see

/turbo-lifecycle

操作	命令
列出数据集	`goldsky dataset list` ⚠️ 速度慢（30-60秒）
验证（必填）	`goldsky turbo validate pipeline.yaml` ✓ 速度快（3秒）
部署/更新管道	`goldsky turbo apply pipeline.yaml`
部署并检查	`goldsky turbo apply pipeline.yaml -i`
列出管道	`goldsky turbo list`
查看实时数据	`goldsky turbo inspect <name>`
检查节点	`goldsky turbo inspect <name> -n <node>`
查看日志	`goldsky turbo logs <name>`
实时跟踪日志	`goldsky turbo logs <name> --follow`
列出密钥	`goldsky secret list`

如需暂停、恢复、重启和删除命令，请查看

/turbo-lifecycle

。

Configuration Reference

配置参考

Pipeline Structure

管道结构

Every Turbo pipeline YAML has this structure:

yaml

name: my-pipeline # Required: unique identifier
resource_size: s # Required: s, m, or l
description: "Optional desc" # Optional: what the pipeline does

sources:
  source_name: # Define data inputs
    type: dataset
    # ... source config

transforms: # Optional: process data
  transform_name:
    type: sql
    # ... transform config

sinks:
  sink_name: # Define data outputs
    type: postgres
    # ... sink config

每个Turbo管道YAML都遵循以下结构：

yaml

name: my-pipeline # 必填：唯一标识符
resource_size: s # 必填：s、m或l
description: "可选描述" # 可选：管道的功能说明

sources:
  source_name: # 定义数据输入
    type: dataset
    # ... 数据源配置

transforms: # 可选：数据处理
  transform_name:
    type: sql
    # ... 转换配置

sinks:
  sink_name: # 定义数据输出
    type: postgres
    # ... 接收器配置

Top-Level Fields

顶层字段

Field	Required	Description
`name`	Yes	Unique pipeline identifier (lowercase, hyphens)
`resource_size`	Yes	Worker allocation: `s` , `m` , or `l`
`description`	No	Human-readable description
`job`	No	`true` for one-time batch jobs (default: `false` = streaming)
`sources`	Yes	Data input definitions
`transforms`	No	Data processing definitions
`sinks`	Yes	Data output definitions

字段	是否必填	描述
`name`	是	唯一管道标识符（小写，仅允许连字符）
`resource_size`	是	工作节点资源分配： `s` 、 `m` 或 `l`
`description`	否	人类可读的管道描述
`job`	否	设置为 `true` 表示一次性批处理任务（默认： `false` = 流式处理）
`sources`	是	数据输入定义
`transforms`	否	数据处理定义
`sinks`	是	数据输出定义

Job Mode

任务模式

Set

job: true

for one-time batch processing (historical backfills, data exports):

yaml

name: backfill-usdc-history
resource_size: l
job: true

sources:
  logs:
    type: dataset
    dataset_name: ethereum.raw_logs
    version: 1.0.0
    start_at: earliest
    end_block: 19000000
    filter: >-
      address = '0xa0b86991c6218b36c1d19d4a2e9eb0ce3606eb48'
transforms: {}
sinks:
  output:
    type: s3_sink
    from: logs
    endpoint: https://s3.amazonaws.com
    bucket: my-backfill-bucket
    prefix: usdc/
    secret_name: MY_S3

Job mode rules:

Runs to completion and auto-cleans up ~1 hour after finishing
Must
goldsky turbo delete
before redeploying — cannot update in-place
Cannot use
restart
— use delete + apply instead
Use
```
end_block
```
to bound the range (otherwise processes to chain tip and stops)
Best with
```
resource_size: l
```
for faster backfills

For architecture guidance on when to use job vs streaming mode, see
/turbo-architecture
.

设置

job: true

用于一次性批处理（历史数据回填、数据导出）：

yaml

name: backfill-usdc-history
resource_size: l
job: true

sources:
  logs:
    type: dataset
    dataset_name: ethereum.raw_logs
    version: 1.0.0
    start_at: earliest
    end_block: 19000000
    filter: >-
      address = '0xa0b86991c6218b36c1d19d4a2e9eb0ce3606eb48'
transforms: {}
sinks:
  output:
    type: s3_sink
    from: logs
    endpoint: https://s3.amazonaws.com
    bucket: my-backfill-bucket
    prefix: usdc/
    secret_name: MY_S3

任务模式规则：

运行完成后，会在约1小时后自动清理资源
重新部署前必须执行
goldsky turbo delete
——不支持原地更新
不支持
restart
命令——请先删除再重新部署
使用
```
end_block
```
来限制处理范围（否则会处理到最新区块后停止）
建议使用
```
resource_size: l
```
以加快回填速度

如需了解何时使用任务模式 vs 流式模式的架构指导，请查看
/turbo-architecture
。

Resource Sizes

资源规格

Size	Workers	Use Case
`s`	1	Testing, low-volume data
`m`	2	Production, moderate volume
`l`	4	High-volume, multi-chain pipelines

规格	工作节点数量	使用场景
`s`	1	测试、低数据量场景
`m`	2	生产环境、中等数据量场景
`l`	4	高数据量、跨链管道场景

Source Configuration

数据源配置

Dataset Source

数据集数据源

yaml

sources:
  my_source:
    type: dataset
    dataset_name: <chain>.<dataset_type>
    version: <version>
    start_at: latest | earliest # EVM chains
    # OR
    start_block: <slot_number> # Solana only

yaml

sources:
  my_source:
    type: dataset
    dataset_name: <chain>.<dataset_type>
    version: <version>
    start_at: latest | earliest # EVM链适用
    # 或者
    start_block: <slot_number> # 仅Solana适用

Source Fields

数据源字段

Field	Required	Description
`type`	Yes	`dataset` for blockchain data
`dataset_name`	Yes	Format: `<chain>.<dataset_type>`
`version`	Yes	Dataset version (e.g., `1.2.0` )
`start_at`	EVM	`latest` or `earliest`
`start_block`	Solana	Specific slot number (omit for latest)
`end_block`	No	Stop processing at this block (for bounded backfills)
`filter`	No	SQL WHERE clause to pre-filter at source level (efficient)

字段	是否必填	描述
`type`	是	对于区块链数据，值为 `dataset`
`dataset_name`	是	格式： `<chain>.<dataset_type>`
`version`	是	数据集版本（例如： `1.2.0` ）
`start_at`	EVM链必填	`latest` 或 `earliest`
`start_block`	Solana必填	特定插槽号（省略则从最新区块开始）
`end_block`	否	处理到该区块后停止（用于限定范围的回填任务）
`filter`	否	SQL WHERE子句，用于在数据源层面预过滤数据（效率更高）

Source-Level Filtering

数据源层面过滤

Use

filter

to reduce data volume before it reaches transforms. This is significantly more efficient than filtering in SQL transforms because it eliminates data at the ingestion layer:

yaml

sources:
  usdc_logs:
    type: dataset
    dataset_name: base.raw_logs
    version: 1.0.0
    start_at: earliest
    filter: >-
      address = lower('0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913')
      AND block_number >= 10000000

Best practices:

Use
```
filter
```
for contract addresses and block ranges (coarse pre-filtering)
Use transform
```
WHERE
```
for event types, parameter values, exclusions (fine-grained)
```
filter
```
uses standard SQL WHERE syntax (same as DataFusion)
Combine
```
filter
```
with
```
start_at: earliest
```
+
```
end_block
```
for precise bounded backfills

使用

filter

在数据到达转换层之前减少数据量，这比在SQL转换中过滤效率高得多，因为它在数据摄入阶段就剔除了不必要的数据：

yaml

sources:
  usdc_logs:
    type: dataset
    dataset_name: base.raw_logs
    version: 1.0.0
    start_at: earliest
    filter: >-
      address = lower('0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913')
      AND block_number >= 10000000

最佳实践：

使用
```
filter
```
过滤合约地址和区块范围（粗粒度预过滤）
使用转换中的
```
WHERE
```
子句过滤事件类型、参数值、排除项（细粒度过滤）
```
filter
```
使用标准SQL WHERE语法（与DataFusion一致）
结合
```
filter
```
与
```
start_at: earliest
```
+
```
end_block
```
实现精准的限定范围回填

Chains and Dataset Types

链与数据集类型

For the full list of chains, prefixes, and dataset types, see

/datasets

. Key points:

EVM chains:

ethereum

base

matic

(Polygon — not

polygon

arbitrum

optimism

bsc

avalanche

Non-EVM:

solana

(uses

start_block

not

start_at

bitcoin.raw

stellar_mainnet

sui

near

starknet

fogo

EVM dataset types:

raw_logs

raw_transactions

(not

transactions

blocks

raw_traces

erc20_transfers

erc721_transfers

decoded_logs

如需完整的链列表、前缀及数据集类型，请查看

/datasets

。重点说明：

EVM链：
```
ethereum
```
、
```
base
```
、
```
matic
```
（Polygon——不要用
```
polygon
```
）、
```
arbitrum
```
、
```
optimism
```
、
```
bsc
```
、
```
avalanche
```

非EVM链：

solana

（使用

start_block

而非

start_at

）、

bitcoin.raw

、

stellar_mainnet

、

sui

、

near

、

starknet

、

fogo

EVM数据集类型：

raw_logs

、

raw_transactions

（不是

transactions

）、

blocks

、

raw_traces

、

erc20_transfers

、

erc721_transfers

、

decoded_logs

Transform Configuration

转换配置

Transform Types

转换类型

Type	Use Case
`sql`	Filtering, projections, SQL functions
`script`	Custom TypeScript/WASM logic
`handler`	Call external HTTP APIs to enrich data
`dynamic_table`	Lookup tables backed by a database

类型	使用场景
`sql`	过滤、投影、SQL函数处理
`script`	自定义TypeScript/WASM逻辑
`handler`	调用外部HTTP API丰富数据
`dynamic_table`	基于数据库的查找表

SQL Transform

SQL转换

Most common transform type:

yaml

transforms:
  filtered:
    type: sql
    primary_key: id
    sql: |
      SELECT
        id,
        sender,
        recipient,
        amount
      FROM source_name
      WHERE amount > 1000

Field	Required	Description
`type`	Yes	`sql`
`primary_key`	Yes	Column for uniqueness/ordering
`sql`	Yes	SQL query (reference sources by name)
`from`	No	Override default source (for chaining)

最常用的转换类型：

yaml

transforms:
  filtered:
    type: sql
    primary_key: id
    sql: |
      SELECT
        id,
        sender,
        recipient,
        amount
      FROM source_name
      WHERE amount > 1000

字段	是否必填	描述
`type`	是	值为 `sql`
`primary_key`	是	用于唯一性/排序的列
`sql`	是	SQL查询（通过名称引用数据源）
`from`	否	覆盖默认数据源（用于链式转换）

TypeScript Transform

TypeScript转换

For complex logic that SQL can't handle (runs in WASM sandbox):

yaml

transforms:
  custom:
    type: script
    primary_key: id
    language: typescript
    from: source_name
    schema:
      id: string
      sender: string
      amount: string
      processed_at: string
    script: |
      function invoke(data) {
        if (data.amount < 1000) return null;  // Filter out
        return {
          id: data.id,
          sender: data.sender,
          amount: data.amount,
          processed_at: new Date().toISOString()
        };
      }

For full TypeScript transform documentation, schema types, and examples, see
/turbo-transforms
.

用于SQL无法处理的复杂逻辑（在WASM沙箱中运行）：

yaml

transforms:
  custom:
    type: script
    primary_key: id
    language: typescript
    from: source_name
    schema:
      id: string
      sender: string
      amount: string
      processed_at: string
    script: |
      function invoke(data) {
        if (data.amount < 1000) return null;  // 过滤掉
        return {
          id: data.id,
          sender: data.sender,
          amount: data.amount,
          processed_at: new Date().toISOString()
        };
      }

如需完整的TypeScript转换文档、 schema类型及示例，请查看
/turbo-transforms
。

Dynamic Table Transform

动态表转换

Updatable lookup tables for runtime filtering (allowlists, blocklists, enrichment):

yaml

transforms:
  tracked_wallets:
    type: dynamic_table
    backend_type: Postgres        # or: InMemory
    backend_entity_name: tracked_wallets
    secret_name: MY_DB            # required for Postgres

Use with

dynamic_table_check()

in SQL transforms:

sql

WHERE dynamic_table_check('tracked_wallets', sender)

For full dynamic table documentation, backend options, and examples, see
/turbo-transforms
.

用于运行时过滤的可更新查找表（白名单、黑名单、数据丰富）：

yaml

transforms:
  tracked_wallets:
    type: dynamic_table
    backend_type: Postgres        # 或：InMemory
    backend_entity_name: tracked_wallets
    secret_name: MY_DB            # 使用Postgres时必填

在SQL转换中结合

dynamic_table_check()

使用：

sql

WHERE dynamic_table_check('tracked_wallets', sender)

如需完整的动态表文档、后端选项及示例，请查看
/turbo-transforms
。

Handler Transform

Handler转换

Call external HTTP APIs to enrich data:

yaml

transforms:
  enriched:
    type: handler
    primary_key: id
    from: my_source
    url: https://my-api.example.com/enrich
    headers:
      Authorization: Bearer my-token
    batch_size: 100
    timeout_ms: 5000

For full handler transform documentation, see
/turbo-transforms
.

调用外部HTTP API丰富数据：

yaml

transforms:
  enriched:
    type: handler
    primary_key: id
    from: my_source
    url: https://my-api.example.com/enrich
    headers:
      Authorization: Bearer my-token
    batch_size: 100
    timeout_ms: 5000

如需完整的Handler转换文档，请查看
/turbo-transforms
。

Transform Chaining

链式转换

Chain transforms using

from

yaml

transforms:
  step1:
    type: sql
    primary_key: id
    sql: SELECT * FROM source WHERE amount > 100

  step2:
    type: sql
    primary_key: id
    from: step1
    sql: SELECT *, 'processed' as status FROM step1

使用

from

实现链式转换：

yaml

transforms:
  step1:
    type: sql
    primary_key: id
    sql: SELECT * FROM source WHERE amount > 100

  step2:
    type: sql
    primary_key: id
    from: step1
    sql: SELECT *, 'processed' as status FROM step1

Sink Configuration

接收器配置

Common Sink Fields

通用接收器字段

Field	Required	Description
`type`	Yes	Sink type
`from`	Yes	Source or transform to read from
`secret_name`	Varies	Secret for credentials (most sinks)
`primary_key`	Varies	Column for upserts (database sinks)

字段	是否必填	描述
`type`	是	接收器类型
`from`	是	读取数据的数据源或转换节点
`secret_name`	视情况而定	凭证密钥（大多数接收器需要）
`primary_key`	视情况而定	用于更新的列（数据库接收器）

Blackhole Sink (Testing)

Blackhole接收器（测试用）

yaml

sinks:
  test_output:
    type: blackhole
    from: my_transform

yaml

sinks:
  test_output:
    type: blackhole
    from: my_transform

PostgreSQL Sink

PostgreSQL接收器

yaml

sinks:
  postgres_output:
    type: postgres
    from: my_transform
    schema: public
    table: my_table
    secret_name: MY_POSTGRES_SECRET
    primary_key: id

Secret format: PostgreSQL connection string:

postgres://username:password@host:port/database

yaml

sinks:
  postgres_output:
    type: postgres
    from: my_transform
    schema: public
    table: my_table
    secret_name: MY_POSTGRES_SECRET
    primary_key: id

密钥格式： PostgreSQL连接字符串：

postgres://username:password@host:port/database

PostgreSQL Aggregate Sink

PostgreSQL聚合接收器

Real-time aggregations in PostgreSQL using database triggers. Data flows into a landing table, and a trigger maintains aggregated values in a separate table.

yaml

sinks:
  balances:
    type: postgres_aggregate
    from: transfers
    schema: public
    landing_table: transfer_log
    agg_table: account_balances
    primary_key: transfer_id
    secret_name: MY_POSTGRES
    group_by:
      account:
        type: text
    aggregate:
      balance:
        from: amount
        fn: sum

Supported aggregation functions:

sum

count

avg

min

max

使用数据库触发器在PostgreSQL中实现实时聚合。数据流入落地表，触发器在单独的表中维护聚合值。

yaml

sinks:
  balances:
    type: postgres_aggregate
    from: transfers
    schema: public
    landing_table: transfer_log
    agg_table: account_balances
    primary_key: transfer_id
    secret_name: MY_POSTGRES
    group_by:
      account:
        type: text
    aggregate:
      balance:
        from: amount
        fn: sum

支持的聚合函数：

sum

、

count

、

avg

、

min

、

max

ClickHouse Sink

ClickHouse接收器

yaml

sinks:
  clickhouse_output:
    type: clickhouse
    from: my_transform
    table: my_table
    secret_name: MY_CLICKHOUSE_SECRET
    primary_key: id

Secret format: ClickHouse connection string:

https://username:password@host:port/database

yaml

sinks:
  clickhouse_output:
    type: clickhouse
    from: my_transform
    table: my_table
    secret_name: MY_CLICKHOUSE_SECRET
    primary_key: id

密钥格式： ClickHouse连接字符串：

https://username:password@host:port/database

Kafka Sink

Kafka接收器

yaml

sinks:
  kafka_output:
    type: kafka
    from: my_transform
    topic: my-topic
    topic_partitions: 10
    data_format: avro          # or: json
    schema_registry_url: http://schema-registry:8081  # required for avro

yaml

sinks:
  kafka_output:
    type: kafka
    from: my_transform
    topic: my-topic
    topic_partitions: 10
    data_format: avro          # 或：json
    schema_registry_url: http://schema-registry:8081  # 使用avro时必填

Webhook Sink

Webhook接收器

Note: Turbo webhook sinks do not support Goldsky's native secrets management. Include auth headers directly in the pipeline config.

yaml

sinks:
  webhook_output:
    type: webhook
    from: my_transform
    url: https://api.example.com/webhook
    one_row_per_request: true
    headers:
      Authorization: Bearer your-token
      Content-Type: application/json

注意：Turbo webhook接收器不支持Goldsky原生密钥管理。请在管道配置中直接包含认证头。

yaml

sinks:
  webhook_output:
    type: webhook
    from: my_transform
    url: https://api.example.com/webhook
    one_row_per_request: true
    headers:
      Authorization: Bearer your-token
      Content-Type: application/json

S3 Sink

S3接收器

yaml

sinks:
  s3_output:
    type: s3_sink
    from: my_transform
    endpoint: https://s3.amazonaws.com
    bucket: my-bucket
    prefix: data/
    secret_name: MY_S3_SECRET

Secret format:

access_key_id:secret_access_key

(or

access_key_id:secret_access_key:session_token

for temporary credentials)

yaml

sinks:
  s3_output:
    type: s3_sink
    from: my_transform
    endpoint: https://s3.amazonaws.com
    bucket: my-bucket
    prefix: data/
    secret_name: MY_S3_SECRET

密钥格式：

access_key_id:secret_access_key

（临时凭证格式：

access_key_id:secret_access_key:session_token

）

S2 Sink

S2接收器

Publish to S2.dev streams — a serverless alternative to Kafka.

yaml

sinks:
  s2_output:
    type: s2_sink
    from: my_transform
    access_token: your_access_token
    basin: your-basin-name
    stream: your-stream-name

发布到S2.dev流——Kafka的无服务器替代方案。

yaml

sinks:
  s2_output:
    type: s2_sink
    from: my_transform
    access_token: your_access_token
    basin: your-basin-name
    stream: your-stream-name

Starter Templates

入门模板

Template files are available in the
templates/
folder. Copy and customize these for your pipelines.

Template	Description	Use Case
`minimal-erc20-blackhole.yaml`	Simplest pipeline, no credentials	Quick testing
`filtered-transfers-sql.yaml`	Filter by contract address	USDC, specific tokens
`postgres-output.yaml`	Write to PostgreSQL	Production data storage
`multi-chain-pipeline.yaml`	Combine multiple chains	Cross-chain analytics
`solana-transfers.yaml`	Solana SPL tokens	Non-EVM chains
`multi-sink-pipeline.yaml`	Multiple outputs	Archive + alerts + streaming

To use a template:

bash

undefined

模板文件位于
templates/
文件夹中。复制并自定义这些模板以创建你的管道。

模板名称	描述	使用场景
`minimal-erc20-blackhole.yaml`	最简单的管道，无需凭证	快速测试
`filtered-transfers-sql.yaml`	按合约地址过滤	USDC、特定代币场景
`postgres-output.yaml`	写入PostgreSQL	生产环境数据存储
`multi-chain-pipeline.yaml`	合并多条链的数据	跨链分析
`solana-transfers.yaml`	Solana SPL代币	非EVM链场景
`multi-sink-pipeline.yaml`	多输出节点	归档+告警+流处理

使用模板的步骤：

bash

undefined

Copy template to your project

复制模板到你的项目

cp templates/minimal-erc20-blackhole.yaml my-pipeline.yaml

Customize as needed, then validate

根据需要自定义，然后验证

goldsky turbo validate my-pipeline.yaml

Deploy

部署

goldsky turbo apply my-pipeline.yaml -i


**Template location:** `templates/` (relative to this skill's directory)

---

goldsky turbo apply my-pipeline.yaml -i


**模板位置：** `templates/`（相对于本技能的目录）

---

Common Update Patterns

常见更新模式

Adding a SQL Transform

添加SQL转换

Before:

yaml

transforms: {}
sinks:
  output:
    type: blackhole
    from: transfers

After:

yaml

transforms:
  filtered:
    type: sql
    primary_key: id
    sql: |
      SELECT * FROM transfers WHERE amount > 1000000
sinks:
  output:
    type: blackhole
    from: filtered # Changed from 'transfers'

更新前：

yaml

transforms: {}
sinks:
  output:
    type: blackhole
    from: transfers

更新后：

yaml

transforms:
  filtered:
    type: sql
    primary_key: id
    sql: |
      SELECT * FROM transfers WHERE amount > 1000000
sinks:
  output:
    type: blackhole
    from: filtered # 从'transfers'改为'filtered'

Adding a PostgreSQL Sink

添加PostgreSQL接收器

yaml

sinks:
  existing_sink:
    type: blackhole
    from: my_transform
  # Add new sink
  postgres_output:
    type: postgres
    from: my_transform
    schema: public
    table: my_data
    secret_name: MY_POSTGRES_SECRET
    primary_key: id

yaml

sinks:
  existing_sink:
    type: blackhole
    from: my_transform
  # 添加新接收器
  postgres_output:
    type: postgres
    from: my_transform
    schema: public
    table: my_data
    secret_name: MY_POSTGRES_SECRET
    primary_key: id

Changing Resource Size

修改资源规格

yaml

resource_size: m # was: s

yaml

resource_size: m # 之前是：s

Adding a New Source

添加新数据源

yaml

sources:
  eth_transfers:
    type: dataset
    dataset_name: ethereum.erc20_transfers
    version: 1.0.0
    start_at: latest
  # Add new source
  base_transfers:
    type: dataset
    dataset_name: base.erc20_transfers
    version: 1.2.0
    start_at: latest

yaml

sources:
  eth_transfers:
    type: dataset
    dataset_name: ethereum.erc20_transfers
    version: 1.0.0
    start_at: latest
  # 添加新数据源
  base_transfers:
    type: dataset
    dataset_name: base.erc20_transfers
    version: 1.2.0
    start_at: latest

Checkpoint Behavior

检查点行为

Understanding Checkpoints

理解检查点

When you update a pipeline:

Checkpoints are preserved by default - Processing continues from where it left off
Source checkpoints are tied to source names - Renaming a source resets its checkpoint
Pipeline checkpoints are tied to pipeline names - Renaming the pipeline resets all checkpoints

当你更新管道时：

默认保留检查点——处理从上次中断的位置继续
数据源检查点与数据源名称绑定——重命名数据源会重置其检查点
管道检查点与管道名称绑定——重命名管道会重置所有检查点

Resetting Checkpoints

重置检查点

Option 1: Rename the source

yaml

sources:
  transfers_v2: # Changed from 'transfers'
    type: dataset
    dataset_name: base.erc20_transfers
    version: 1.2.0
    start_at: earliest # Will process from beginning

Option 2: Rename the pipeline

yaml

name: my-pipeline-v2 # Changed from 'my-pipeline'

Warning: Resetting checkpoints means reprocessing all historical data.

选项1：重命名数据源

yaml

sources:
  transfers_v2: # 从'transfers'修改而来
    type: dataset
    dataset_name: base.erc20_transfers
    version: 1.2.0
    start_at: earliest # 将从最开始处理

选项2：重命名管道

yaml

name: my-pipeline-v2 # 从'my-pipeline'修改而来

警告： 重置检查点意味着需要重新处理所有历史数据。

Troubleshooting

故障排查

See

references/troubleshooting.md

for:

CLI hanging / Turbo binary not found fixes
Common validation errors (unknown dataset, missing primary_key, bad source reference)
Common runtime errors (auth failed, connection refused, Neon size limit)
Quick troubleshooting table

Also see

/turbo-monitor-debug

for error patterns and log analysis.

请查看

references/troubleshooting.md

获取以下内容：

CLI卡住/Turbo二进制文件未找到的解决方法
常见验证错误（未知数据集、缺失primary_key、无效数据源引用）
常见运行时错误（认证失败、连接拒绝、Neon大小限制）
快速故障排查表

如需错误模式和日志分析，请查看

/turbo-monitor-debug

。