Logstash 필터와 파이프라인

Welcome to an advanced look at Logstash! So far, you've learned how to get logs into Logstash and apply basic filters. But what happens when your data gets more complex?

In this lesson, we'll explore powerful techniques like conditional logic, managing multiple pipelines, and leveraging advanced filters for intricate data transformations. This will help you handle real-world logging challenges.

Not all logs are created equal! You might have different log formats coming from various services, or you might want to process events differently based on their content.

Conditional logic allows Logstash to apply filters or outputs only when certain conditions are met. This is achieved using if statements, similar to programming languages.

• Use if to check field values, tags, or other event properties.
• Apply specific filters or actions only to matching events.

Let's see a simple example. We'll use if to check if a field named type exists and has a specific value. If it does, we'll add a tag.

Try inputting {"message": "Hello", "type": "app_log"} and then {"message": "World"} to see the difference.

Just like in programming, you can extend your conditional logic with else if and else blocks. This allows for more complex branching and ensures every event is handled appropriately.

Logstash executes these conditions sequentially. The first matching condition's block is executed, and then it moves on.

• if [field] == "value": Executes if the condition is true.
• else if [another_field] == "another_value": Executes if the first if was false, and this condition is true.
• else: Executes if none of the preceding if or else if conditions were true.

As your system grows, you might be collecting logs from many different sources (e.g., web servers, databases, security devices). Each source might require entirely different processing logic.

Multiple pipelines allow you to isolate and manage these distinct processing flows independently. Instead of one giant, complex Logstash configuration, you can have several smaller, focused ones.

• Isolation: Errors in one pipeline won't affect others.
• Resource Management: Assign specific resources to different pipelines.
• Modularity: Easier to develop, test, and maintain configurations.

To use multiple pipelines, you define them in a file called pipelines.yml, usually located in your Logstash configuration directory (e.g., /etc/logstash/pipelines.yml).

Each entry specifies a unique ID, the path to its configuration file (.conf), and optional settings like number of worker threads.

Example pipelines.yml:

Sometimes, built-in Logstash filters aren't enough for very specific or complex data transformations. That's where the ruby filter comes in!

The ruby filter allows you to execute arbitrary Ruby code within your Logstash pipeline. This provides immense flexibility to manipulate events in ways not possible with standard filters.

• Use for: Complex string manipulations, mathematical operations, custom data lookups, or logic that depends on multiple fields.
• Caution: Can impact performance if not used carefully.

Let's use the ruby filter to create a new field that combines parts of existing fields and calculates a value.

Try inputting: {"user_id": "123", "item_count": 5, "price_per_item": 10.5}

The mutate filter is a workhorse for basic field manipulation, but it has some advanced operations that are incredibly useful:

• split: Splits a string field into an array based on a delimiter.
• join: Joins an array field into a string using a specified separator.
• convert: Changes the data type of a field (e.g., string to integer, float to string).
• rename: Changes the name of an existing field.

These operations help you shape your data precisely for storage and analysis.

Great job! You've leveled up your Logstash skills. We covered:

• How conditional logic (if, else if, else) allows for dynamic event processing.
• The benefits and configuration of multiple pipelines for modular and isolated data flows.
• Leveraging the powerful ruby filter for highly custom data transformations.
• Advanced operations within the mutate filter like split, join, and convert.

These techniques are crucial for building robust and adaptable Logstash configurations for complex, real-world data.