Operations | Monitoring | ITSM | DevOps | Cloud

If there’s one thing every IT leader can agree on, it’s this: Manual Active Directory (AD) management never ends. There’s always one more access request, one more approval chain, and one more audit reminder flashing on your screen. By the time you’ve closed your last ticket of the day, there’s already another one waiting. For many teams, 2025 became the year of “we’ll automate next quarter.” But next quarter came and went without any automation.

We are thrilled to announce a major update to our Python Connectors line. The release adds support for Python 3.14, PostgreSQL 18, MySQL 9, and MariaDB 12, introduces modern authentication and security options, and delivers notable performance gains across several connectors.

As organizations rely on increasingly complex IT infrastructures, incident management often turns into a constant cycle of alerts, escalations, and fixes. While reactive responses may keep operations running, they rarely address the deeper systemic issues that slowly erode performance. Recurring incidents, silent failures, and hidden patterns are usually symptoms of unresolved root causes that traditional approaches struggle to uncover.

Traditional UI testing struggles to keep up with rapid design and workflow changes, often focusing on brittle selectors rather than user outcomes. Harness AI Test Automation introduces intent-driven, natural language assertions that understand what teams want to verify, not just how tests are written.

A feast of new features. A cornucopia of new capabilities. A banquet of breakthroughs (and the T-day puns are just getting started). Graylog 7.0 brings a full plate of advancements that help security teams cut through noise, control cloud costs, and respond with confidence. We’re serving practical improvements across dashboards, automation, and AI support so analysts can focus on action instead of manual effort.

Like many of you, over the last couple of years, I’ve been using AI, or, well, let’s just name it appropriately, Large Language Models (LLM), as a part of my job. I’ve also used it in my hobby. With it, I’ve generated snippets of code, tested data conversions, even built a small database for a presentation. However, to date, I haven’t tried doing everything through the LLM. Now, I’m going to.

GraphQL isn’t just another API protocol—it’s a new layer of abstraction. It collapsed dozens of REST endpoints into one flexible interface where clients decide what data to fetch and how deep to go. That freedom is a gift for front-end teams and a headache for anyone tasked with reliability. Traditional monitoring doesn’t work here. A REST endpoint can be pinged for uptime.

In 2022, we introduced Grafana Mimir, our open source, horizontally scalable, multi-tenant time series database (TSDB) designed for long-term storage of Prometheus and OpenTelemetry metrics. Over the years, Mimir has become a go-to metrics backend within the open source community, with 30 project maintainers and more than 4.7k GitHub stars.

IT infrastructure is vast, complex, and interdependent. At any point in time, businesses rely on thousands of servers running thousands of processes. Detecting server downtime is fairly easy—but true observability is when you know precisely which processes are working as intended and which are silently contributing to performance degradation. A failed database worker or a memory-leaking background service can silently drain resources until your most critical apps grind to a halt.

We're excited to announce one of our most frequently requested features: inverted DNS checks. This new capability allows you to detect when DNS records are created (whether they're planned infrastructure changes or rogue records) and prevent unwanted values from appearing in your DNS configuration.