AWS CloudWatch

What is Amazon CloudWatch?

Amazon CloudWatch is a monitoring tool used by DevOps engineers, developers, project managers, and IT managers to provide data insights and what to do with your applications, infrastructure, and AWS resources. You collect all the data you need in metrics and logs, and you can access it in a single place. It comes with features like alarms and events data to help make the process successful and give developers more time to improve the business aspect of applications.

Amazon CloudWatch can do a lot of work with the data, like letting the users know the health of their applications and resources, troubleshooting errors, performing analysis, etc. After collecting data, it aggregates it and summarises useful information like computing resources, i.e., memory, disk size etc.

Take your career to next level in AWS with HKR. Enroll now to get Aws Online Training

How Amazon Cloudwatch works?

Amazon Cloudwatch uses metrics in a repository. All the metrics come from different applications and AWS components like Amazon EC2. When a developer needs statistics, you need to visit the repository.

After calculating the metrics' statistics, you submit the data to the CloudWatch console. Amazon CloudWatch Alarm has different actions that you can take when the criteria you set get met. Some actions include starting, stopping, or terminating the Amazon EC2 instances. It also provides an option of auto-scaling using Amazon Simple Notification Service (Amazon SNS) actions.

It has data centres found in specific regions to avoid failure cases and increase stability.

How to Access CloudWatch

There are several methods of accessing CloudWatch: These methods are

(i) AWS CLI (Command Line Interface)

• Download the AWS CLI from https://awscli.amazonaws.com/AWSCLIV2.ms and double-click the executable file to run it.
• Confirm the installation of AWS CLI by typing aws --version on the command line.
• Use the specific commands that work with AWS CloudWatch to access it.

(ii) AWS SDK
The SDK supports different programming languages that you can use to access CloudWatch.The commonly used language is Java. To use the SDK, you follow the following steps:

• Log in to your AWS account and grab your access keys and other AWS credentials.
• Download the Java SDK from https://sdk-for-java.amazonwebservices.com/latest/aws-java-sdk.zip and copy it to your IDE directory like Eclipse.
• Install the SDK and configure the environment.
• Run different commands to access the CloudWatch.Amazon has an SDK guide you can look at and practice.

(iii)Amazon CloudWatch console

• Access the CloudWatch console by typing the following URL in your browser  https://console.aws.amazon.com/cloudwatch/.
• Navigate to the region part and locate the region you would like to use to host your application.
• Create an alarm by navigating to your Alarms, choosing the Create Alarm, and following the prompts to finish the process.
• All your metrics will show, or you can enable them if it's not showing yet as some need configuration.

(iv) CloudWatch API 
You can carry out API requests using HTTP or HTTPS. When carrying out the queries, don't worry about which one is right.CloudWatch supports all queries from POST, GET, DELETE and PUT requests, and they have the same result. When querying, you use several parameters like Action or Operation. To understand more about CloudWatch API, comprehensive documentation covers more about the API.

Amazon CloudWatch Events

It uses simple rules when setting up events and routing them with streams. It enables developers to track real-time events that happen when using AWS services.CloudWatch Events notifies the user of any operational changes, and it takes action by using different ways like collecting state information, sending messages to the environment, etcc.
CloudWatch Events can sometimes use cron expectations to schedule different actions. Some of the AWS services that work well with CloudWatch Events include Amazon EC2 instances, CodeBuild projects, Amazon ECS tasks, Amazon SQS queues, AWS Batch jobs, etc.

When using CloudWatch Events, there are several terminologies you have to understand:

(i)Events
An event is when there is a change in the AWS environment. Most AWS resources generate events when there are state changes. Users can generate custom events and later publish them to CloudWatch Events. Users can also schedule these events to happen after a certain period.

(ii)Rules
Rules help match the coming events by routing them to their target for processing. They mostly get processed randomly, which helps different parts of a company focus on events they like or want. One rule has the option of routing several targets. One of the common methods is using a JSON file that you can customize and send to the target.

(iii)Targets
They are for processing events. Most of the targets are AWS services, i.e., AWS Lambda functions, Amazon SNS topics, Amazon ECS tasks, Amazon EC2 instances, Amazon SQS queues, etc. Most of the events are in JSON formats. We also have built-in targets that can process events.

AWS Training

• Master Your Craft
• Lifetime LMS & Faculty Access
• 24/7 online expert support
• Real-world & Project Based Learning

Explore Curriculum

CloudWatch Metrics

Metrics measure the performance of the applications and systems using data. Most AWS services produce free metrics, and you can use them to monitor some of the resources. Amazon CloudWatch uses the metrics to analyse and create alarms and graphs after loading them into your account.

Metrics are useful, and you can access them for up to 15 months, both historical and real-time data either per second or minute. You can use the CloudWatch Metrics Insights tool to create graphs that can help in the identification of patterns and trends.

CloudWatch dashboards

Dashboards are pages that you can use to view and monitor all the resources in one place. It gives users the power to customise them to meet their needs by changing the metrics and alarms of the AWS components. It helps company teams understand what to do when there is a suspicious incident or issue.

Teams can also use the application measurements and resources view to improve communications when performing operational events. You can check the health of the application resources either in one or multiple regions by using single views of metrics and alarms. Users can use display graphs or different colours when tracking multiple regions.

Creating a dashboard is easier. You can use the PutDashboard API or the AWS console or CLI. To access the dashboard, one must consider the three policies: AdministratorAccess, CloudWatchFullAccess, and custom policy.

If you have any doubts on AWS, then get them clarified from AWS Industry experts on our AWS Community! 

What Are CloudWatch Logs?

CloudWatch logs help users monitor the log files from Amazon components like Amazon EC2 instance, Route 53, Cloudtrail, etc. It helps you put all the logs in one place. It gives you an easy option of viewing them, checking any error codes, filtering the logs, and archiving them for future use.

You collect all the logs from applications, AWS components and services, and systems. It helps you create dashboards using customised queries and visualisations. Some of the other features of CloudWatch logs include:

• It uses insights to query and analyse log data that helps in responding to any issues faster.
• You can use it to track any errors and monitor all the applications using log data.
• It supports the use of alarms to notify the events, helping monitor the logged events.
• It keeps the logs for a long time, up to 10 years.
• It helps you archive the log data and use it only when needed.
• It provides all the information about the DNS queries.

Features of Amazon CloudWatch

Some of the features of AWS CloudWatch include the following:

(i) Collection and storage of logs

• It has a Log service that helps in the collection and storage of logs from different applications, AWS components, and other services. There are several types of logs collected, and they have different sources. These are:
• Vended logs-they come from AWS, and they get published on behalf of the users. Examples are Amazon Route 53 and Amazon VPC Flow Logs.
• Logs by AWS services that supports over 20 AWS services. Examples are AWS CloudTrail, AWS Lambda, and Amazon API Gateway.
• Custom logs are the ones you get from your applications.
• It has PutMetricData API and CloudWatch Agent that helps in the publication of metrics into CloudWatch.It also has built-in metrics for the detection of errors. It collects application metrics like memory usage, errors, and user activities from applications to help troubleshoot and monitor trends and performance. One can collect all the infrastructure metrics from over the 50 AWS components like AWS Lambda, Amazon S3, Amazon EC2, Amazon API Gateway etc.
• Container Insights collects metrics from container logs and performance metrics from resources like CPU, disk information, memory, network, etc. Later, it aggregates them. It also supports the collection of logs from applications and AWS services like Amazon EKS/k8s data plane, Amazon EC2 instance etc.
• It has CloudWatch Lambda Insights that make the collection of logs from AWS Lambda functions easier. It also collects other metrics from performance criteria like memory while generating other custom metrics to monitor and alarm processes.
• It has Amazon CloudWatch Metric Streams that help in streaming of metrics easier by sending them to third parties using HTTP endpoints. It also fastens the process of streaming metrics to your preferred destination, like data lakes.

(ii) Security

AWS CloudWatch uses AWS Identity and Access Management (IAM), which limits the users who can access certain data. It also has AWS Key Management Service (AWS KMS) that helps in the encryption of log groups to provide more compliance and security.

(iii) Monitoring

• It has Amazon CloudWatch RUM that enables developers to have a better view of the performance of the application and on the client's side. It also helps in reducing the meantime to resolution.
• It has Amazon CloudWatch ServiceLens that helps in the visualisation and analysis of the application's performance and health in one place.
• It uses alarms for setting threshold metrics for resources.
• It has Anomaly Detection that uses machine learning technology to analyse metrics data that helps in the detection of abnormal behaviour.
• It has Amazon CloudWatch Synthetics that checks and monitors application endpoints 24/7 and ensures they are working well.

(iv) Act

• It supports auto-scaling due to alarms that notify developers of the threshold usage. It helps in managing the resources and having better planning.
• Container Insights ensures the alarm provides notification on the computing metrics, triggering some policies to stop, reboot, or recover the Amazon EC2 instances.
• CloudWatch Events tracks real-time system events when there are any changes to the AWS components. It enables developers to respond faster and take appropriate action.

(v) Analysing

• Amazon CloudWatch Logs Insights provides more intelligence about your logs, checking any operational issues on servers and software.
• It has Contributor Insights that works with time-series data to provide contributors that can affect the system's performance.
• It has Amazon CloudWatch, Evidently, enabling developers to perform different experiments to identify errors when implementing new applications.

Top 30 frequently asked AWS Devops Interview Questions !

Subscribe to our youtube channel to get new updates..!

Subscribe

Benefits of CloudWatch

• It uses a single platform for all the analytics. When using CloudWatch, applications release data that consists of logs, metrics, and events. It makes it easier to use the data since it organises all the data from AWS services and applications in one place, making it easier to break down the data and come up with solutions quicker in case of any problem.
• It provides troubleshooting insights using the logs. Cloudwatch has CloudWatch Logs Insights, which gives users the chance to view and analyse all the logs. They charge for the queries you will run. You can also improve the visibility by creating the metrics for the logs, creating alarms, and creating a correlation between logs and the metrics.
• It improves performance and reduces resource usage. Users can use machine learning technology to check if there are abnormalities in the logs and metrics. It has Cloudwatch events that help you work with other AWS services like AWS Lambda, AWS CloudFormation, and Amazon SNS. Users can also pause the instances to reduce the billing.
• It's easy to collect metrics from AWS and its products. Cloudwatch works well with other AWS services like Lambda, DynamoDB, Amazon S3, Amazon EC2, etc. It makes monitoring all the AWS resources and applications easy. It provides CloudWatch API that helps in monitoring the AWS services in hybrid environments. Users can automatically set the metrics to get published after a short period to help you have a deeper look at the logs.
• It provides insights and operational visibility.CloudWatch has dashboards where users can view data within a one-second range that goes up to 15 months. You can also perform math on the metrics to check any operational insights. It helps in ensuring the resources get utilised and optimization of performance.

Amazon CloudWatch pricing

It uses the pay as you use model, and it gets billed at the end of the month. It has two payment plans: the free tier and the Paid plan.

(I) Free tier

You use CloudWatch for free and enjoy other AWS services like Amazon Kinesis, Amazon EC2, and Amazon S3. Some of the features you enjoy include:

• Monitoring metrics that comprise ten detailed monitoring at 1-minute frequency, basic monitoring at 5-minute frequency, and one million API requests except for the GetMetric ones.
• Access to all events with custom events excluded.
• 3 dashboards that run up to 50 metrics per month.
• 10 alarm metrics.
• RUM of 1 million events per month
• Synthetics of 100 canary runs per month.

(ii) Paid plan

Paid tier uses different factors to come with the billing. Amazon provides an AWS pricing calculator to help you find the estimated costs. Some of the features it considers include

• Number of dashboards
• Alarms
• The number of metrics per month for both the API and streams.
• Logs.
• Contributor insights.
• Canaries.
• Events.

Some of the billing examples are:

For a more detailed pricing model, Amazon provides a comprehensive pricing guide.

Challenges of AWS CloudWatch

Some of the challenges one can face when using AWS CloudWatch are:

• The pricing is a little higher when compared with other tools.
• It only supports integration when used with other AWS components.
• Its dashboard doesn't support other analytics visuals like histograms.
• It requires technical knowledge to use, making its users only skilled individuals and teams.

AWS CloudWatch Use Cases

• It handles the monitoring and troubleshooting of infrastructure. It provides logs and metrics for applications and infrastructure, the creation of alarms making it easy to diagnose any performance issues within the AWS premises.
• It provides good analytics. When analysing the operational logs and metrics, you can identify new patterns in the application performance. You can use queries to have visualisations that can detect issues quickly.
• Monitoring applications. It collects data from the application infrastructure and the front end. It helps the user monitor the applications that run on the AWS premises. It also has ServiceLens, which checks the app's performance and detects issues using the metrics, logs, and traces. It also has canaries that monitor all the UI workflows and application endpoints.
• It optimises resources. It uses alarms to monitor the threshold of the resource you specified. If the alarms start, it makes the Amazon EC2 autoscale or stops the instance to do another resource planning.
• It increases the MTTR(mean time to resolution). When you have metrics and logs, you can use them together with data from AWS X-Ray to improve the observability. You can also use the user requests to make the troubleshooting process faster and reduce the meantime to resolution.

Want to know more about AWS , visit here AWS Tutorial !

Conclusion

We now have a deeper understanding of AWS Cloudwatch and its features. Most developers who use AWS CloudWatch to monitor their applications are happy due to its automation. No manual setup is needed when deploying apps and using the AWS services. It has made it easier to get the errors and correct them quickly. It works well with the infrastructure and hybrid environments. If you desire to get a job in DevOps and Cloud technology field, this is one of the tools you can learn.

• What is AMI in AWS
• What is VPC in AWS