E-Handbook: Smart strategies for application monitoring in the cloud Article 4 of 4

WavebreakmediaMicro - Fotolia


Assess native cloud monitoring tools from top IaaS providers

Monitoring tools from Google, AWS and Azure all share a common goal: help developers optimize application performance. But they also have key differences that users should note.

Before enterprises choose a cloud provider, they need to ensure its native management and monitoring tools fit their needs.

AWS CloudWatch, Azure Monitor and Google Strackdriver are the most popular native cloud monitoring tools. They all provide basic features to capture data, maintain lognd use metrics and dashboards to analyze application performance. Many popular third-party cloud monitoring tools also use data from these native tools under the covers.

However, there are distinctions between these three tools. For example, AWS CloudWatch is strongest in autoscaling applications, because of its integration with other AWS offerings, such as CloudFormation. It also provides a high level of customization. But Microsoft and Google have stronger sets of features around troubleshooting application performance issues. In addition, Google Stackdriver can capture data from AWS, and Azure Monitor has capabilities to capture data from AWS and Google Cloud.

Here's a breakdown of some other key features, and differences, between these three native cloud monitoring tools.

AWS CloudWatch

The most mature monitoring and logging tool among the bunch, Amazon CloudWatch gathers metrics and provides granularity via customizable dashboards and alerts. It enables developers to capture a wide range of metrics from events and resources.

With custom dashboard support, developers can also easily set up different views of infrastructure and applications within a particular context. For example, developers can create one view for all the components of a particular business application to identify the source of a performance problem.

Amazon CloudWatch can gather data from AWS IaaS services, such as EC2 and S3, as well as from SaaS applications and private enterprise infrastructure. While it does not work seamlessly with other cloud platforms, such as Azure or Google, CloudWatch does support an API to integrate with other analytics tools.

In addition to keeping track of application performance, CloudWatch supports a rich ecosystem of ancillary AWS tools for autoscaling and application resiliency. However, it does not directly support application troubleshooting, which is instead provided by other AWS tools, such as X-Ray. Another limitation of AWS CloudWatch is its lack of code profiling capabilities to correlate performance problems with specific lines of code.

Azure Monitor

Azure Monitor is a service that consolidates two former Microsoft cloud monitoring tools -- Log Analytics and Application Insights.

The Azure Monitor dashboard, which includes several starter templates, enables cloud admins to visualize and organize metrics and log data from Azure resources. There are also features, such as diagnostic logs, which make it easier to analyze different infrastructure components, including an application gateway or load balancer. The tool supports distributed tracing and code profiling.

Enterprises can configure alerts to autoscale infrastructure in response to CPU usage or application performance thresholds. Azure Monitor also includes a health section, which makes it easier to correlate application performance issues with information about Azure as a whole.

However, Azure Monitor is currently limited in its ability to correlate the performance of serverless Azure Functions with overall application performance.

Google Stackdriver

Like the other native cloud monitoring tools mentioned above, Google Stackdriver includes a rich set of features to capture logs, event data and other metrics. While it can also retrieve data from AWS infrastructure, this integration is primarily for the purpose of diagnostics and does not provide the same autoscaling capabilities as AWS CloudWatch.

Google Stackdriver provides a set of tools for debugging, distributed tracing and code profiling. The Stackdriver Profiler tool enables developers to correlate code with application performance issues. This only works when developers store application code in a Google Cloud Source Repository, GitHub or Bitbucket repository, but it does streamline the process of debugging live applications and inspecting application calls and local variables for apps that run exclusively on Google Cloud Platform.

Dig Deeper on Cloud app development and management

Data Center