سجلات CloudWatch ومجموعات السجلات

Imagine trying to fix a problem in your application without knowing what happened! That's where logs come in.

CloudWatch Logs is an AWS service that helps you collect, monitor, and store logs from your applications and various AWS services.

It's crucial for troubleshooting, auditing, and understanding how your systems are performing.

At the core of CloudWatch Logs is the concept of a Log Group.

Think of a Log Group as a dedicated folder or container for logs that share the same characteristics, like logs from a specific application or service.

For example, all logs from your 'WebApp-Production' service might go into one Log Group.

Inside each Log Group, you'll find Log Streams.

A Log Stream is a sequence of log events from a single source within that Log Group.

• For an EC2 instance, each instance might have its own Log Stream.
• For a Lambda function, each invocation or version could generate a Log Stream.

They help you pinpoint the exact source of your log data.

Many AWS services automatically create Log Groups for you when they generate logs, like AWS Lambda or AWS Fargate.

For custom applications, you'll typically create a Log Group manually and then configure your application or the CloudWatch Agent to send logs to it.

You can create and manage Log Groups via the AWS Management Console, AWS CLI, or SDKs.

To send logs from applications running on an EC2 instance to CloudWatch Logs, you use the CloudWatch Agent.

• Install the agent on your EC2 instance.
• Configure it to monitor specific log files (e.g., /var/log/myapp.log).
• The agent then streams these logs to a specified Log Group and Log Stream in CloudWatch Logs.

AWS Lambda automatically integrates with CloudWatch Logs! Any output your Lambda function generates (e.g., using print() in Python or console.log() in Node.js) is sent to CloudWatch Logs.

Try running this simple Python Lambda function and check its logs:

Once your logs are in CloudWatch Logs, you can easily search and filter them.

• Search: Look for specific keywords, phrases, or patterns across all your log events.
• Filter: Narrow down results by time range, Log Stream, or even specific fields if your logs are structured (e.g., JSON).

This helps you quickly diagnose issues or find relevant events.

Logs can consume a lot of storage, and you might have compliance requirements for how long to keep them.

Log Retention Policies allow you to define how long CloudWatch Logs should store your log events.

• Options range from 'Never Expire' to 1 day, 30 days, 1 year, etc.
• Setting appropriate retention helps manage costs and adhere to data policies.

Beyond just viewing logs, you can extract meaningful data from them using Metric Filters.

A Metric Filter defines a pattern to look for in your log events (e.g., the word "ERROR" or a specific status code).

When the pattern is matched, CloudWatch increments a custom metric, which you can then use to create alarms (as seen in the previous lesson!).

Time for a quick check on CloudWatch Logs!

Which statement best describes the relationship between CloudWatch Log Groups and Log Streams?

Great job! You've learned the essentials of CloudWatch Logs.

We covered:

• Log Groups for organizing logs.
• Log Streams for individual log sources.
• How to ingest logs from services like EC2 and Lambda.
• How to search and filter logs efficiently.
• Configuring retention policies and creating metric filters from logs.

Next, we'll explore CloudTrail for event-driven monitoring and auditing.