Operations | Monitoring | ITSM | DevOps | Cloud

In the August episode of This Month in Datadog, Jeremy shares how you can make more informed cloud cost decisions, gain insights into your LiteLLM-powered applications, and secure Kubernetes infrastructure with Datadog Workload Protection. Later in the episode, Danny puts the spotlight on Datadog Kubernetes Autoscaling, which helps you deliver cost savings without sacrificing performance.

There are a lot of ways to gauge the performance of your DevOps teams and the health of your software, but DORA metrics have emerged as the industry standard. If you aren’t familiar with DORA metrics, take a few minutes to read this comprehensive guide to understanding DORA metrics. DORA metrics were designed to offer a high-level, long-term view of how your teams are performing.

When we started building Logs in Sentry we had one goal: make them useful for real debugging, not just another high-volume text storage. This meant making them "trace connected" from day one. This let us ensure they were tightly connected to the actions and performance happening in your application, right where developers already go to investigate errors, performance, and latency issues. Now, Logs is out of beta and generally available to everyone.

APM tracing records the complete execution path of a request as it travels through your system, including database queries, external API calls, cache lookups, message queue events, and inter-service requests. Each step is captured with precise start and end timestamps, duration, and context such as service name, operation name, and relevant attributes. This lets you pinpoint where latency or errors originate without piecing together metrics and logs manually.

A bit more behind-the-scenes work than usual this month, but I still managed to ship some public-facing features you might be interested in. Logging in, and clicking around the dashboard just got 60% faster, and we're just getting started.

The most challenging question in AI today isn’t how to build with it. It’s whether you can prove it’s worth what you’re spending on it. Every week, I hear the same thing from engineering and finance leaders: “We know the AI bills are big.

If your SaaS organization is experimenting with OpenAI, your cloud bill just got a new line item. And unless you know exactly what drives it, that line item can go from manageable to margin-killer, fast. It’s also worth clarifying that OpenAI is not the same as ChatGPT. ChatGPT is the familiar end-user app with a flat monthly subscription. OpenAI, meanwhile, is the platform behind it — a mix of models, features, and usage-based pricing that shifts depending on what you build.

Your application is only as reliable as the infrastructure it runs on. Most commonly, that means Kubernetes is doing the job by managing fleets of containers, scaling services on demand, and keeping workloads distributed across nodes. Traditional dashboards weren’t built to scale with this reality. They give you snapshots of raw metrics. They don’t scale to multi-cluster environments. They don’t map relationships between resources.

Serverless architecture has changed how applications are built and run. It removes the need to manage servers, letting developers focus on writing code while automatically scaling with demand. But even with its pay-as-you-go model, serverless apps can get expensive if not monitored and optimized. In this blog, lets see how smart serverless monitoring helps developers and DevOps engineers lower serverless costs, boost performance, and keep operations running smoothly.

Modern PHP applications rarely exist in isolation. They run across distributed environments, connect to MySQL or PostgreSQL databases, interact with Redis or Memcached, rely on APIs, and communicate with microservices. This interconnected web brings power but also enormous complexity. When performance issues arise, finding the root cause can feel like searching for a needle in a haystack. Is it the database? A caching layer? A failing third-party API?