tencent cloud

masukan

Creating Processing Task

Terakhir diperbarui:2024-01-20 17:44:35

    Overview

    The data processing feature provides the capabilities to filter, cleanse, mask, enrich, and distribute log data. It can be benchmarked against the open-source component Logstash, but its input and output are CLS log topics.
    
    
    
    You can select a source log topic and write Domain Specific Language (DSL) processing function statements to process logs line by line and send logs that meet certain requirements to specified target log topics.If your business uses only processed logs, we recommend you configure the log retention period of the source log topic as one day and not enable the index in order to reduce costs.

    Billing

    Data processing incurs fees. For example, if you have a log topic in Guangzhou and 100 GB raw data that becomes 50 GB after being filtered, and the target topic is written (with index not enabled), you need to pay:
    Data processing fees: 100 GB * 0.026 USD/GB = 2.6 USD
    Write traffic fees of the target topic: 50 GB * 0.3 (compression ratio) * 0.032 USD/GB = 0.48 USD
    Storage fees of the target topic: 50 GB * 0.3 (compression ratio) * 0.0024 USD/GB = 0.036 USD
    The total is 3.116 USD, which is generally the case for regions in the Chinese mainland. For billing details in special regions, see Product Pricing.

    Use cases

    For most users, data processing can achieve the following:
    Extract structured data to facilitate future operations such as search, analysis, and dashboard generation.
    Simplify logs to reduce use costs. Unnecessary logs will be discarded to reduce storage and traffic costs.
    Mask sensitive data, such as users' ID numbers and phone numbers.
    Ship logs by category. For example, logs are categorized by ERROR, WARNING, or INFO and then distributed to different log topics.

    Common DSL functions

    The following describes some DSL functions that are commonly used for data processing. The corresponding use cases can be viewed in DSL Statement Generator on the Create Data Processing Task > Edit Processing Statement page in the console. You can directly copy the sample code and modify parameters as needed for quick use.
    Extracting structured data: ext_sep() for extracting by separator, ext_json or ext_json_jmes for extracting by JSON, and ext_regex() for extracting by regular expression
    Simplifying logs: log_drop for discarding a line of log and fields_drop() for discarding log fields
    Masking sensitive data: regex_replace()
    Classifying logs: t_if_else() and t_switch() for determining conditions, and log_output() for distributing logs
    Editing fields: fields_set() for adding/resetting fields, and fields_rename() for renaming fields
    Determining the presence of a field or data in logs: has_field() for determining whether a field exists and regex_match() for determining whether certain data exists
    Comparing numerical values in logs: op_gt() (greater than), op_ge() (greater than or equal to), and op_eq() (equal to); functions for addition, subtraction, multiplication, and division, such as op_mul() for multiplication and op_add() for addition
    Processing text in logs: str_uppercase() for uppercase or lowercase switching and str_replace() for text replacement
    Processing time values in logs: dt_to_timestamp() for converting a time value to a UTC timestamp
    The preceding are common DSL functions used in data processing scenarios. The DSL statement generator also provides other DSL functions for your reference. You can copy a desired function to the DSL function editing box, modify parameters, and use it.
    The following describes how to create a data processing task.

    Prerequisites

    You have activated CLS, created a source log topic, and successfully collected log data.
    You have created a target log topic, which is recommended to be empty to allow the writing of processed data.
    The current account has the permission to configure data processing tasks.
    Note:
    Data processing can process only real-time log streams but not historical logs.
    If the data processing console does not automatically load raw log data, the possible cause is that your log topic does not have real-time log streams. In that case, you need to manually add custom logs in JSON format to complete writing the data processing script.

    Directions

    1. Log in to the CLS console.
    2. On the left sidebar, click Data Processing.
    3. Click Create Processing Task.
    
    
    4. On the Basic Info page, configure the following information:
    
    
    Task Name: Enter the custom task name.
    Source Log Topic: Select the source log topic.
    Target Log Topic: Enter the target name and select the log topic. Here, the target name can be the alias of the target log topic and used as an input parameter of the written DSL function. You can configure up to ten target log topics.
    5. Click Next.
    6. On the Edit Processing Statement page, perform the following operations:
    A DSL function can be tested by two types of data: raw log data and custom data. The system automatically loads raw log data, 100 records by default. If you think the raw log data is insufficient for testing the DSL function, you can directly enter custom data on the Custom Data tab. Alternatively, you can click Add Custom Data on the Raw Data tab, modify the data, and use it as your custom data.
    
    
    Note:
    Custom data must be in JSON format.
    If there are multiple entries of custom data, add them in the following format:
    [
    {
    "content": "2021-10-07 13: 32: 21|100|Customer not checked in|Jack|Beijing|101123199001230123|"
    },
    {
    "content": "2021-10-07 13: 32: 21|500|Checked in successfully|Jane|Sanya|501123199001230123|"
    },
    {
    "content": "2021-10-07 13: 32: 21|1000|Checked out|Lily|Sanya|101123196001230123|"
    }
    ]
    7. Click DSL Statement Generator.
    8. In the pop-up window, select a function, and click Insert Function. The DSL statement generator provides the descriptions and examples of multiple types of functions. You can copy and paste the examples to the processing statement editing box and modify parameters as needed to write your own DSL functions. You can also refer to Processing Example to quickly understand how to write a DSL function.
    9. After writing the DSL processing statement, click Preview or Checkpoint Debugging to run and debug the DSL function. The running result will be displayed at the bottom right of the page. You can adjust the DSL statement according to the running result until it meets your requirements.
    
    
    10. Click OK to submit the data processing task.
    Hubungi Kami

    Hubungi tim penjualan atau penasihat bisnis kami untuk membantu bisnis Anda.

    Dukungan Teknis

    Buka tiket jika Anda mencari bantuan lebih lanjut. Tiket kami tersedia 7x24.

    Dukungan Telepon 7x24