Logs

Logs provide detailed records of events and activities in your AI applications.

Overview

The Logs collection provides a way to collect, store, and analyze logs from your AI applications. Logs can:

Record detailed information about events and activities
Help diagnose issues and troubleshoot problems
Provide insights into application behavior
Support audit and compliance requirements

Key Features

Collection: Collect logs from various sources
Storage: Store logs securely and efficiently
Search: Search and filter logs
Analysis: Analyze logs for patterns and insights

Log Types

Observability.do supports various log types:

Application Logs

Record application events and activities:


// Example application log
{
  timestamp: '2023-06-30T12:34:56.789Z',
  level: 'info',
  service: 'user-service',
  environment: 'production',
  message: 'User login successful',
  context: {
    userId: 'user-123',
    loginMethod: 'email',
    ipAddress: '192.168.1.1'
  }
}

Model Logs

Record model interactions and outputs:


// Example model log
{
  timestamp: '2023-06-30T12:35:10.123Z',
  level: 'info',
  service: 'content-generation',
  environment: 'production',
  message: 'Model generated content',
  context: {
    modelId: 'gpt-4',
    promptTokens: 150,
    completionTokens: 250,
    totalTokens: 400,
    latency: 1250,
    requestId: 'req-456'
  }
}

System Logs

Record system events and activities:


// Example system log
{
  timestamp: '2023-06-30T12:36:20.456Z',
  level: 'warn',
  service: 'api-gateway',
  environment: 'production',
  message: 'High CPU usage detected',
  context: {
    cpuUsage: 85,
    memoryUsage: 70,
    diskUsage: 60,
    instanceId: 'i-789'
  }
}

Audit Logs

Record security and compliance events:


// Example audit log
{
  timestamp: '2023-06-30T12:37:30.789Z',
  level: 'info',
  service: 'auth-service',
  environment: 'production',
  message: 'User permission changed',
  context: {
    userId: 'user-123',
    adminId: 'admin-456',
    permission: 'data-access',
    action: 'grant',
    reason: 'Project assignment'
  }
}

Logging Levels

Observability.do supports standard logging levels:

trace: Detailed information, typically useful only for diagnosing problems
debug: Information useful for debugging
info: General information about system operation
warn: Warning events that might cause problems
error: Error events that might still allow the application to continue running
fatal: Severe error events that will likely lead the application to abort

Logging in Code

Add logging to your code using the Observability.do API:


// Import the logger
import { logger } from '@drivly/observability'
 
// Configure the logger
logger.configure({
  service: 'user-service',
  environment: 'production',
  defaultContext: {
    version: '1.2.0',
  },
})
 
// Log at different levels
logger.trace('Detailed trace information', { detail: 'value' })
logger.debug('Debug information', { debugKey: 'value' })
logger.info('User login successful', { userId: 'user-123' })
logger.warn('Rate limit approaching', { currentRate: 95, limit: 100 })
logger.error('Failed to process request', { requestId: 'req-456', error: 'Database connection failed' })
logger.fatal('Service is shutting down', { reason: 'Out of memory' })
 
// Log with additional context
logger.info('User profile updated', {
  userId: 'user-123',
  fields: ['name', 'email'],
  source: 'api',
})
 
// Create a child logger with additional context
const userLogger = logger.child({ userId: 'user-123' })
userLogger.info('User login successful') // Automatically includes userId
 
// Log function execution
const result = await logger.withLogging(
  async () => {
    // Function implementation...
    return { success: true }
  },
  {
    message: 'Processing user request',
    level: 'info',
    context: { requestId: 'req-456' },
    logResult: true,
  },
)

Querying Logs

Query logs using the Observability.do API:


// Search logs
const logs = await observability.logs.search({
  query: 'level:error AND service:user-service',
  timeRange: {
    start: '2023-06-01T00:00:00Z',
    end: '2023-06-30T23:59:59Z',
  },
  limit: 100,
  offset: 0,
  sort: {
    field: 'timestamp',
    order: 'desc',
  },
})
 
// Get logs for a specific request
const requestLogs = await observability.logs.getByRequestId('req-456')
 
// Get logs for a specific user
const userLogs = await observability.logs.search({
  query: 'context.userId:user-123',
  timeRange: {
    start: '2023-06-01T00:00:00Z',
    end: '2023-06-30T23:59:59Z',
  },
})
 
// Analyze log patterns
const errorPatterns = await observability.logs.analyze({
  query: 'level:error',
  timeRange: {
    start: '2023-06-01T00:00:00Z',
    end: '2023-06-30T23:59:59Z',
  },
  groupBy: ['service', 'context.error'],
  metrics: ['count'],
})

Log Management

Manage your logs through the dashboard or API:


// Configure log retention
await observability.logs.configureRetention({
  default: {
    duration: '30d',
  },
  levels: {
    error: {
      duration: '90d',
    },
    fatal: {
      duration: '365d',
    },
  },
})
 
// Export logs
const exportResult = await observability.logs.export({
  query: 'level:error AND service:user-service',
  timeRange: {
    start: '2023-06-01T00:00:00Z',
    end: '2023-06-30T23:59:59Z',
  },
  format: 'json',
  destination: {
    type: 's3',
    bucket: 'logs-backup',
    prefix: 'exports/2023-06',
  },
})
 
// Create a log view
await observability.logs.createView({
  name: 'error-logs',
  query: 'level:error',
  description: 'View of all error logs',
})