Announcing Support for Oracle Arm-Based Ampere A1 Instances | Datadog

Announcing support for Oracle Arm-based Ampere A1 instances

Author Kai Xin Tai
Author Jimmy Caputo

Published: May 28, 2021

Arm processors have long been at the center of mobile computing, powering billions of smartphones, tablets, smartwatches, and other IoT devices. Today, these processors are beginning to see broader adoption in the cloud as they promise better performance, higher energy efficiency, and lower costs than their x86-based predecessors. Just this week, Oracle announced its new Oracle Cloud Infrastructure Ampere A1 Compute platform, built on the Ampere Altra Arm processor.

The Datadog Agent for Arm allows you to collect metrics, traces, logs, and more from your Oracle Arm-based Ampere A1 instances (both virtual machine and bare metal), so you can get comprehensive visibility into your entire infrastructure. In this post, we’ll discuss how Datadog can help you:

Monitor migrations to Oracle Arm-based Ampere A1 instances

If you’re migrating existing workloads over to these new Oracle Arm-based Ampere A1 instances, it’s important to compare the health and performance of both your old and new instances to verify that the process has not introduced any regressions. Datadog’s host map provides a bird’s eye view of your entire infrastructure, allowing you to track system-level metrics such as CPU and memory utilization across all of your hosts. You can use any tag you’ve configured to filter and group your hosts for a more organized view of your fleet. For example, if you’re shifting from AMD E3 instances to Ampere A1 instances, you can group your hosts by instance type to easily compare their resource utilization, as shown in the screenshot below.

Use Datadog's host map to compare CPU utilization across instances

Rightsizing your instances to fit your workloads

Whether you’re running a small database or a multi-tier web service, Oracle lets you flexibly configure the number of OCPUs and amount of memory on your Ampere A1 instances to suit your workload requirements. But because these resources are billed at a per-second granularity, you’ll want to optimize your instance configuration for both performance and cost.

Oracle allows you to configure OCPU and memory on its Ampere A1 instances
Configuring the OCPU and memory allocation of an Ampere A1 instance in the Oracle Cloud Infrastructure Console

Datadog APM collects request traces and performance metrics such as throughput, error rate, and latency, so you can better understand how your applications are responding to incoming traffic. You can also view this data in context with system-level metrics to determine whether your instances' available resources can support your workloads. This enables you to identify and proactively scale up any resource-constrained instances to avoid potential performance degradations. Similarly, you can use APM alongside the host map to identify over-provisioned instances, which can be scaled down to save costs.

Use Datadog APM to monitor service-level metrics like request count, latency, and errors

Optimize resource-intensive parts of your application

While Datadog’s host maps and infrastructure metrics can alert you to resource saturation, Live Processes allows you to pinpoint the most resource-intensive processes, which may be preventing other processes from running efficiently. In Live Processes, you can dive into every process across your distributed system in one place, without having to SSH into each individual host. Datadog also lets you visualize processes running on a host in a tree format (similar to what you’d get from running htop), enabling you to easily spot orphan processes that can be terminated. Additionally, you can generate metrics from these processes to track and analyze trends in resource consumption over the long term.

Live Processes lets you view processes across all your hosts to pinpoint

Datadog Continuous Profiler goes one step further by providing granular, code-level insights. As shown in the screenshot below, each profile contains a collection of stack traces that reveal which methods or packages in your code are consuming the most resources. And because Continuous Profiler collects data from all of your hosts without interruptions, you can track how code modifications affect performance over time.

A profile that visualize which methods in your code are taking up the most CPU time

Start monitoring Oracle Arm-based Ampere A1 instances with Datadog today

Datadog’s support for Oracle Arm-based Ampere A1 instances extends your infrastructure coverage and allows you to monitor all your virtual machines, on-premise servers, containers, and more in one place. And with our 450+ out-of-the-box integrations, including Oracle Database and MySQL, you can see how the performance of your infrastructure impacts the services running on it.

If you’re an existing Datadog customer, you can deploy the Datadog Agent for Arm on your Ampere A1 instances right away. You can also read our joint blog post with Oracle for more useful resources. New to Datadog? Get started with a 14-day today.