Skip to main content

​
Log Data Integration with ELK

Below is the technical implementation of Logs.
1

Add “Logger Util” created to your project and import into your program.

​
Python

from logger import api_logger or import logging

​
Java

    import com.locai.logger.config.LoggerConfig;
    import com.locai.logger.config.LoggerFactory;
    import com.locai.logger.model.LogEvent;
or import org.slf4j.Logger; // for direct object usage
2

Initialize the 'logger_sdk' with service and environment details

Follow lower case for Environment, servicename, Service Name: {productname}–{be/ai/fe}–{environment}.

​
Python

 from logger import api_logger

 logger_sdk = LoggerSDK(service_name="-{product-fe/be/..}-Environment -logs", # Required: Service identifier  ex: development-iqraa-fe
     environment="development",                   # Required: Environment (dev/stage/prod)
     product_name="your-product/project-name",    # Required: Project identifier
     app_name="your-app-name",                    # Required: Name of your application ex: user-registration-api
     log_dir="/path/to/logs",                     # Required: Directory for log files
     tenant_id="tenant_id"                        # Required: Default tenant ID
 )

​
Java

LoggerSDK logger = new LoggerSDK( " Environment-{product-fe/be/..}-logs ",   // Required: Service name ex: development-iqraa-fe
    "development",           // Required: Environment
    "your-product-name",     // Required: Product name EX: iqraa
    "your-app-name",          // Required: Application name
    "/path/to/logs",         // Required: Log directory
    "default-tenant"         // Optional: Default tenant ID
);
3

Basic Logging

  • Use “Context” for adding the details related to the function(like API, method, backend data), and runtime details etc.
  • Extra entity to store supplementary or additional information specific to the event or operation being logged.like task completion, or process, user specific detalis etc..

    • ​
    Python

    logger_sdk.log( level="INFO", // Required: Log level (INFO/ERROR/WARN)
event_name="user_action",              //  Required: Event identifier
message="User performed action",   //  Required: Log message
user_id="user123",      //  Required: User identifier
tenant_id="tenant456",    // Required: Tenant identifier
extra={            //  Optional: Additional data
    "action_type": "login",
    "ip_address": "192.168.1.1"
 }
)

    ​
    Java

    logger.log( "info",  // level Required: Log level (info/error/debug/warn)
    "user_action",                   // eventName Required
    "User performed action",         // message Required
    null,// context (Map<String, Object>)
    Map.of("action_type", "login"), // extra data Optional
    null,         // durationMs Optional
    "user123",  // userId  Required
    "tenant456"   // tenantId Required
);
    4

    Decorator for logging function duration

    ​
    Python

    @logger_sdk.log_duration("create_user")     # Optional:
def create_user(user_data):
# Your function code here
Pass

    ​
    Java

    String userId = "user123";
String tenantId = "tenant456";

UserResponse result = logger.logDuration(
"create_user",
userId,
tenantId,
() -> {
// Your operation here
return createUser(userData);
}
);
    5

    Error Logging

    Incase if the application deployed using kubernaties, and logs are logged into standard output so below steps and log directory paths may not necessary for now. Also, you may need to enable configuration for to write logs to stdout/stderr.

    ​
    Python

    try:
# Your code here
raise Exception("Something went wrong")
except Exception as e:
logger_sdk.log(
level="error",
event_name="error_event",
message=str(e),
extra={"traceback": traceback.format_exc()}
)

    ​
    Java

    try {
// Your code here
throw new Exception("Something went wrong");
} catch (Exception e) {
logger.log(
    "error", "error_event", e.getMessage(),
    null, Map.of("stackTrace", e.getStackTrace()),
    null, userId, tenantId);
}
    Reference Notes:
  • Added the log structure which can be used common for API/Backend/database by passing relavent details like for context, extra sub-entities.
  • Added log structures specific for API/Backend/database.