Skip to content

Latest commit

 

History

History
474 lines (359 loc) · 16.7 KB

README_USER.md

File metadata and controls

474 lines (359 loc) · 16.7 KB
Raw

go-etl Data Synchronization User Manual

go-etl's datax is a data synchronization tool that currently supports data synchronization between mainstream relational databases such as MySQL, Postgres, Oracle, SQL Server, DB2, and file formats like CSV and XLSX.

1 Where to Download

You can download the 64-bit version of the binary program for Windows or Linux operating systems from the latest release.

2 Getting Started

Obtain the datax binary program from the downloaded package. On Linux, as shown in the Makefile, export LD_LIBRARY_PATH=/home/ibmdb/clidriver/lib. This library can be downloaded from ibm db2, otherwise it will not run.

Additionally, for Oracle, you need to download the corresponding 64-bit version of the ODBC dependency from Oracle.

2.1 Single-Task Data Synchronization

Calling datax is straightforward; you can simply invoke it as follows:

data -c config.json

-c specifies the data source configuration file.

When the return value is 0 and it displays "run success," it indicates successful execution.

When the return value is 1 and it displays "run fail," it provides the reason for the failure.

2.1.1 Data Source Configuration File

The data source configuration file is a JSON file that combines data sources. For example, to synchronize data from MySQL to Postgres:

{
    "core" : {
        "container": {
            "job":{
                "id": 1,
                "sleepInterval":100
            }
        }
    },
    "job":{
        "content":[
            {
                "reader":{
                    "name": "mysqlreader",
                    "parameter": {
                        "username": "test:",
                        "password": "test:",
                        "column": ["*"],
                        "connection":  {
                                "url": "tcp(192.168.15.130:3306)/source?parseTime=false",
                                "table": {
                                    "db":"source",
                                    "name":"type_table"
                                }
                            },
                        "where": ""
                    }
                },
                "writer":{
                    "name": "postgreswriter",
                    "parameter": {
                        "username": "postgres",
                        "password": "123456",
                        "writeMode": "insert",
                        "column": ["*"],
                        "preSql": [],
                        "connection":  {
                                "url": "postgres://192.168.15.130:5432/postgres?sslmode=disable&connect_timeout=2",
                                "table": {
                                    "schema":"destination",
                                    "name":"type_table"
                                }
                         },
                        "batchTimeout": "1s",
                        "batchSize":1000
                    }
                },
               "transformer":[]
            }
        ]
    }
}

The configurations for reader and writer are as follows:

Type Data Source Reader (Read) Writer (Write) Documentation
Relational Database MySQL/Mariadb/Tidb √ √ Read, Write
Postgres/Greenplum √ √ Read, Write
DB2 LUW √ √ Read, Write
SQL Server √ √ Read, Write
Oracle √ √ Read, Write
Unstructured Stream CSV √ √ Read, Write
XLSX (excel) √ √ Read, Write

2.1.2 Usage Examples

Note: On Linux, as indicated in the Makefile, export LD_LIBRARY_PATH=${DB2HOME}/lib

2.1.2.1 Synchronizing with MySQL
  • Initialize the database using cmd/datax/examples/mysql/init.sql for testing purposes
  • Start the MySQL synchronization command:
datax -c examples/mysql/config.json
2.1.2.2 Synchronizing with PostgreSQL
  • Initialize the database using cmd/datax/examples/postgres/init.sql for testing purposes
  • Start the PostgreSQL synchronization command:
datax -c examples/postgres/config.json
2.1.2.3 Synchronizing with DB2
  • Before use, download the corresponding DB2 ODBC library, e.g., make dependencies and release.bat for Linux
  • Note: On Linux, as indicated in the Makefile, export LD_LIBRARY_PATH=${DB2HOME}/lib
  • Note: On Windows, as indicated in release.bat, set path=%path%;%GOPATH%\src\github.com\ibmdb\go_ibm_db\clidriver\bin
  • Initialize the database using cmd/datax/examples/db2/init.sql for testing purposes
  • Start the synchronization command:
datax -c examples/db2/config.json
2.1.2.4 Synchronizing with Oracle
  • Before use, download the corresponding Oracle Instant Client. For example, it is recommended to download version 12.x to connect to Oracle 11g.
  • Note: On Linux, export LD_LIBRARY_PATH=/opt/oracle/instantclient_21_1:$LD_LIBRARY_PATH. Additionally, install libaio.
  • Note: On Windows, set path=%path%;%GOPATH%\oracle\instantclient_21_1. Oracle Instant Client 19 no longer supports Windows 7.
  • Initialize the database using cmd/datax/examples/oracle/init.sql for testing purposes
  • Start the synchronization command:
datax -c examples/oracle/config.json
2.1.2.5 Synchronizing with SQL Server
  • Initialize the database using cmd/datax/examples/sqlserver/init.sql for testing purposes
  • Start the SQL Server synchronization command:
datax -c examples/sqlserver/config.json
2.1.2.6 Synchronizing CSV to PostgreSQL
  • Initialize the database using cmd/datax/examples/csvpostgres/init.sql for testing purposes
  • Start the synchronization command:
datax -c examples/csvpostgres/config.json
2.1.2.7 Synchronizing XLSX to PostgreSQL
  • Initialize the database using cmd/examples/datax/csvpostgres/init.sql for testing purposes (Note: The path may need correction as it seems inconsistent with other examples)
  • Start the synchronization command:
datax -c examples/xlsxpostgres/config.json
2.1.2.8 Synchronizing PostgreSQL to CSV
  • Initialize the database using cmd/datax/examples/csvpostgres/init.sql for testing purposes
  • Start the synchronization command:
datax -c examples/postgrescsv/config.json
2.1.2.9 Synchronizing PostgreSQL to XLSX
  • Initialize the database using cmd/datax/examples/csvpostgres/init.sql for testing purposes (Note: The initialization script may not be specific to XLSX synchronization)
  • Start the synchronization command:
datax -c examples/postgresxlsx/config.json
2.1.2.10 Other Synchronization Examples

In addition to the above examples, all data sources listed in the go-etl features can be used interchangeably. Configurations can be set up for data sources such as MySQL to PostgreSQL, MySQL to Oracle, Oracle to DB2, etc.

2.1.3 Global Database Configuration

{
    "job":{
        "setting":{
            "pool":{
              "maxOpenConns":8,
              "maxIdleConns":8,
              "connMaxIdleTime":"40m",
              "connMaxLifetime":"40m"
            },
            "retry":{
              "type":"ntimes",
              "strategy":{
                "n":3,
                "wait":"1s"
              },
              "ignoreOneByOneError":true
            }
        }
    }
}
2.1.3.1 Connection Pool (pool)
  • maxOpenConns: Maximum number of open connections
  • maxIdleConns: Maximum number of idle connections
  • connMaxIdleTime: Maximum idle time for a connection
  • connMaxLifetime: Maximum lifetime of a connection
2.1.3.2 Retry (retry)

ignoreOneByOneError: Whether to ignore individual retry errors

  • Retry types and strategies:
    1. Type ntimes: Retry a fixed number of times. Strategy: "strategy":{"n":3,"wait":"1s"}, where n is the number of retries and wait is the waiting time between retries.
    2. Type forever: Retry indefinitely. Strategy: "strategy":{"wait":"1s"}, where wait is the waiting time between retries.
    3. Type exponential: Exponential backoff retry. Strategy: "strategy":{"init":"100ms","max":"4s"}, where init is the initial waiting time and max is the maximum waiting time.
2.1.3.3 Test Data
datax -c examples/global/config.json

2.1.4 Using Split Keys

It is assumed that data is evenly distributed based on the split key. Proper use of such a split key can make synchronization faster. Additionally, to speed up queries for maximum and minimum values, preset maximum and minimum values can be used for large tables.

2.1.4.1 Testing Approach
  • Generate MySQL data using a program and create split.csv
cd cmd/datax/examples/split
go run main.go
  • Use init.sql to create the table
  • Synchronize to the MySQL database
cd ../..
datax -c examples/split/csv.json
  • Modify examples/split/config.json to set the split key as id,dt,str
  • Synchronize MySQL data with integer, date, and string types using the split key
datax -c examples/split/config.json

2.1.5 Using preSql and postSql

preSql and postSql are sets of SQL statements executed before and after writing data, respectively.

2.1.5.1 Testing Approach

In this example, a full import is used:

  1. Before writing data, a temporary table is created.
  2. After writing data, the original table is deleted, and the temporary table is renamed to the new table.
datax -c examples/prePostSql/config.json

2.1.6 Flow Control Configuration

Previously, the byte and record configurations for speed did not take effect. Now, with the introduction of flow control, byte and record will be effective. byte limits the size of cached messages in bytes, while record limits the number of cached messages. If byte is set too low, it can cause the cache to be too small, resulting in failed data synchronization. When byte is 0 or negative, the limiter will not work. For example, byte can be set to 10485760, which is equivalent to 10Mb (10x1024x1024).

{
    "job":{
        "setting":{
            "speed":{
                "byte":10485760,
                "record":1024,
                "channel":4
            }
        }
    }
}
2.1.6.1 Flow Control Testing
  • Generate src.csv using a program and initiate the flow control test
cd cmd/datax/examples/limit
go run main.go
cd ../..
datax -c examples/limit/config.json

2.1.7 querySql Configuration

The database reader uses querySql to query the database.2.2 多任务数据同步

  • 2.2.1 Usage

    2.2.1.1 Data Source Configuration File

    Configure the data source configuration file, such as syncing from MySQL to PostgreSQL:

    {
    "core" : {
        "container": {
            "job":{
                "id": 1,
                "sleepInterval":100
            }
        }
    },
    "job":{
        "content":[
            {
                "reader":{
                    "name": "mysqlreader",
                    "parameter": {
                        "username": "test:",
                        "password": "test:",
                        "column": ["*"],
                        "connection":  {
                                "url": "tcp(192.168.15.130:3306)/source?parseTime=false",
                                "table": {
                                    "db":"source",
                                    "name":"type_table"
                                }
                            },
                        "where": ""
                    }
                },
                "writer":{
                    "name": "postgreswriter",
                    "parameter": {
                        "username": "postgres",
                        "password": "123456",
                        "writeMode": "insert",
                        "column": ["*"],
                        "preSql": [],
                        "connection":  {
                                "url": "postgres://192.168.15.130:5432/postgres?sslmode=disable&connect_timeout=2",
                                "table": {
                                    "schema":"destination",
                                    "name":"type_table"
                                }
                         },
                        "batchTimeout": "1s",
                        "batchSize":1000
                    }
                },
               "transformer":[]
            }
        ]
    }
}
    2.2.1.2 Source-Destination Configuration Wizard File

    The source-destination configuration wizard file is a CSV file. Each row of configuration can be set as follows:

    path[table],path[table]

    Each column can be a path or table name. Note that all tables should have a configured database name or schema name, which needs to be configured in the data source configuration file.

    2.2.1.3 Batch Generation of Data Configuration Sets and Execution Scripts
    datax -c tools/testData/xlsx.json -w tools/testData/wizard.csv

    -c specifies the data source configuration file, and -w specifies the source-destination configuration wizard file.

    The execution result will generate a set of configuration files in the data source configuration file directory, with the number of rows in the source-destination configuration wizard file. The configuration sets will be named as specified_data_source_config_file1.json, specified_data_source_config_file2.json, ..., specified_data_source_config_file[n].json.

    Additionally, an execution script named run.bat or run.sh will be generated in the current directory.

    2.2.1.4 Batch Execution of Generated Data Configuration Sets
    Windows
    run.bat

    Linux

    run.sh

    2.2.2 Test Results

    You can run the test data in cmd/datax/testData:

    cd cmd/datax
datax -c testData/xlsx.json -w testData/wizard.csv

    The result will generate a set of configuration files in the testData directory, with the number of rows in the wizard.csv file. The configuration sets will be named as xlsx1.json, xlsx2.json, ..., xlsx[n].json.

    2.3 Data Synchronization Help Manual

    2.3.1 Help Command

    datax -h

    Help display:

    Usage of datax:
  -c string
        config (default "config.json")
  -http string
        http
  -w string
        wizard

    -http adds a listening port, such as 8080. After enabling, access 127.0.0.1:8080/metrics to get real-time throughput.

    2.3.2 View Version

    datax version

    Display: version number (git commit: git commit number) compiled by go version go version number

    v0.1.0 (git commit: c82eb302218f38cd3851df4b425256e93f85160d) compiled by go version go1.16.5 windows/amd64

    2.3.3 Enable Monitoring Port

    datax -http :8443 -c examples/limit/config.json
    2.3.3.1 Get Current Monitoring Data

    Use a web browser to access http://127.0.0.1:8443/metrics to get the current monitoring data:

    {"jobID":1,"metrics":[{"taskGroupID":0,"metrics":[{"taskID":0,"channel":{"totalByte":2461370,"totalRecord":128624,"byte":3820,"record":191}}]}]}
    • totalByte: Total number of bytes synchronized
    • totalRecord: Total number of records synchronized
    • byte: Number of bytes synchronized in the channel
    • record: Number of records synchronized in the channel