The F-Files Podcast

Five columns. One sensitive table. And only a handful of people who see what’s really there. In this episode, Frank and Kevin explore the invisible line between data protection and privileged access in Microsoft Fabric. With dynamic data masking enabled on the EmployeeData table, most users see only obfuscated results — but not everyone.They dive into:How dynamic data masking works across Fabric’s data warehousing layerThe critical difference between SELECT and UNMASK permissionsWhy masking rules are your first — and sometimes last — line of defenseBest practices for protecting fields like SSNs, email addresses, and more🔍 The Mask Falls for One — because visibility isn’t just about roles, it’s about trust.🎧 Tune in and uncover what lies beneath… if you're allowed.

The content was identical. The configurations were not. And the damage? Very real. In this episode, Frank and Kevin uncover the critical importance of maintaining distinct settings between Development, Test, and Production environments in Microsoft Fabric. What looks like a smooth deployment often hides a dangerous assumption: that what works in one stage will work in all.They explore:How to use selective deployment to control what actually moves forwardWhy unique deployment rules per stage are non-negotiableThe danger of assigning one workspace across all environmentsReal-world stories of environments gone rogue due to overlooked configs⚠️ Deploy Divide is a reminder that content consistency means nothing without contextual integrity.🎧 Listen now — before your next deployment crosses a line it shouldn’t.

In this fiery episode of The F Files, Frank and Kevin find themselves deep in the inferno of inefficient pipelines, where static values, hardcoded paths, and inflexible components reign supreme. Data transformations are brittle, errors lurk in every corner, and updates require a full-scale rescue mission.But not all hope is lost.Armed with parameterization, string interpolation, and smart pipeline design, the duo sets out to break free from the shackles of hardcoding. Can reusable pipeline logic and dynamic inputs bring order to chaos? Or will they be trapped forever in a loop of copy-paste despair?Tune in to uncover:- How to use pipeline parameters for flexible, scalable design- The role of string interpolation in dynamic configurations- Real-world examples of escaping hardcoded nightmare scenariosBecause in the world of data engineering…“Hell is other people's hardcoded pipelines.”

The logs showed a deployment to Production. The problem? No one on the team admitted doing it. In this episode, Frank and Kevin investigate the chilling consequences of improper permission control in Microsoft Fabric deployment pipelines. When pipeline settings aren’t locked down, the line between authorized changes and accidental disasters gets dangerously thin.They uncover:- How to assign pipeline admin roles and why it matters- The difference between workspace permissions and pipeline access- Why role-based access control (RBAC) is your only real defense- Real-world horror stories of "harmless" edits with very real consequences🔐 The Unauthorized Push is a cautionary tale about access, accountability — and what happens when you leave the door wide open.🎧 Listen now. Before someone pushes something they shouldn't.

One workspace in Dev. The same workspace in Prod. And somewhere in between… the rules started to bend. In this episode, Frank and Kevin dive into the chaos and complexity of deployment pipelines in Microsoft Fabric — where content transitions between environments are anything but linear. As teams demand selective deployment, backward promotion, and environmental isolation, the system starts to show its flux. They explore:- How to configure multi-stage deployment pipelines for Dev → Test → Prod- Why selective deployment is critical — and what happens when it’s not enabled- The risks of full deployment in a dynamic workspace structure- What it means to maintain stability when environments are in constant motion⚠️ Fabric Flux — because in a world of constant deployment, nothing stays where it was meant to be.🎧 Listen now, before your pipeline starts drifting too.

The deployment completed successfully. No errors. No warnings. Just one overlooked connection string… and two weeks of corrupted reports. In this episode, Frank and Kevin uncover a cautionary tale from deep inside a Microsoft Fabric deployment pipeline. When a production stage is connected to the wrong database, the damage is invisible — until it’s irreversible.They discuss:- Why deployment rules are critical for configuring production resources- How to correctly manage connection strings across Dev, Test, and Prod- When manual overrides lead to silent disasters- And how one deployment triggered a full-blown audit💡 The Pipeline Confession is a story of trust, automation — and what happens when neither is enough.🔍 Listen now. Before your pipeline reveals its own secrets.

The deployment pipeline was ready. Dev passed. Test looked perfect. And then… someone pushed everything to Production. In this episode, Frank and Kevin uncover the risks of reckless deployment in Microsoft Fabric. When you're running an analytics solution across Development, Test, and Production, not all content belongs in the final environment. But how do you control what gets deployed — and what stays behind?They’ll explore:- The role of selective deployment in protecting production environments- Why deployment rules matter — especially between Test and Prod- What happens when one workspace gets pushed too far- Practical strategies to keep your deployment pipeline clean and safe⚠️ Deploy with Caution — because once it's in Production, there's no turning back.🛠️ Listen now before your pipeline surprises you.

In the mysterious world of Microsoft Fabric, not everything is as it seems—especially when it comes to data filtering. Join Frank and Kevin as they uncover the hidden power behind Smart Expressions in Fabric. What looks like a simple conditional clause hides immense potential for optimizing performance, streamlining logic, and customizing analytics at scale.How do Smart Expressions actually work under the hood? What pitfalls should you avoid when building them into your pipelines and semantic models? And why are they becoming a critical tool in the modern data engineer’s arsenal?Tune in for expert insights, real-world use cases, and maybe even a few secrets Microsoft didn’t want you to know...The truth is in the logic.The answer is out there.

In this gripping installment of The F Files, Kevin and Frank return to dissect the aftermath of a hasty deployment in Microsoft Fabric. When rollback isn't an option and temp tables turn traitor, our heroes face the ghost of decisions past.Discover the real-world implications of flawed deployment strategies, missing backups, and the underestimated complexity of production environments. From “it worked in dev” to “why is prod on fire?”, this episode dives into the dangers of cutting corners in Fabric-powered data architectures.💥 Learn how to:- Prevent costly mistakes by validating before you deploy- Implement controlled rollbacks in deployment pipelines- Use versioned artifacts and test environments effectivelySometimes second chances come too late — but you can still avoid Reloaded Regret.

Something strange is happening in the Spark clusters.Memory thresholds are breached. Jobs mutate. Performance vanishes. And in the middle of it all, a single question lingers: Did we configure the pool… or awaken something else?In this episode, Frank and Kevin dig deep into the mysterious world of executor memory management in Microsoft Fabric. What happens when your jobs outgrow the starter pool? Is it really just a configuration issue — or is there something more sinister lurking in your autoscaling?They’ll unpack:- The limits of the starter Spark pool- Why some jobs demand more memory than expected- How to fix performance using custom pools and dynamic allocation- When to say: "This job needs more than Fabric wants to give"🧠 The Memory Mutation is real — and it starts with your next workload.👾 Stream the episode now before your cluster crashes first.