The Bit Bucket

In real-time intelligence, everything starts with the concept of an event. An event is the most fundamental unit of real-time data — it’s simply a record that something happened.

That something could be almost anything, depending on your business. In finance, an event might be a stock trade or a payment transaction. In a web application, it might be a customer clicking a button, logging in, or abandoning a shopping cart. In IoT, it could be a sensor reading like temperature, vibration, or GPS coordinates. Even a server log entry or an error message can be considered an event.

2026-01-19

This is a post in the SQL Interview series. These aren’t trick or gotcha questions, they’re just questions designed to scope out a candidate’s knowledge around SQL Server and Azure SQL Database.

Section: Development Level: Medium

Question:

You have tables that have both constraints and triggers. You need to explain the order of execution for these. Which execution order (from first to last), are used for:

• Check constraints
• After triggers
• Instead of triggers

Answer:

2026-01-18

Microsoft Fabric Real-Time Intelligence is a complete toolkit — an end-to-end set of capabilities that allow you to take streaming data, make sense of it, and act on it.

It starts with event ingestion. Fabric can connect to a wide range of streaming sources: Kafka, Azure Event Hubs, IoT Hub, and many others. That means whether your data is coming from IoT sensors, application logs, or business systems, you can bring it all into Fabric without a lot of custom wiring. Fabric RTI doesn’t force you to pick one source — it’s designed to be open and flexible.

2026-01-17

This is a post in the SQL Interview series. These aren’t trick or gotcha questions, they’re just questions designed to scope out a candidate’s knowledge around SQL Server and Azure SQL Database.

Section: Administration Level: Advanced

Question:

If a database is configured for simple recovery model, what could stop the truncation of the transaction log being effective, and for the transaction log to keep growing in size?

Answer:

Here are the most frequent reasons:

2026-01-16

Batch processing has been the backbone of data analytics for decades. The idea is simple: you collect data over a period of time, maybe hours or a whole day, and then process it in one big chunk. This is how traditional ETL pipelines and overnight data warehouse loads work. It’s efficient when immediacy doesn’t matter — for example, producing a daily sales report each morning.

But the limitation is obvious: if you need to react quickly, batch just doesn’t cut it. By definition, you’re waiting for the batch window to complete before you see the results. If fraud is happening right now, or if a customer is struggling with your app this very minute, a batch report tomorrow morning is far too late.

2026-01-15

This is a post in the SQL Interview series. These aren’t trick or gotcha questions, they’re just questions designed to scope out a candidate’s knowledge around SQL Server and Azure SQL Database.

Section: Administration Level: Intro

Question:

You are restoring a database backup. The restore finished reading pages 10 minutes ago, but the database still shows that it is restoring.

Can you speed up the restoring phase by restarting the server? If so, should you do that? If not, why not?

2026-01-14

When people hear the phrase real-time, their minds often jump straight to sub-second response. They imagine dashboards updating instantly or transactions being analyzed faster than the blink of an eye. But in reality, not every scenario demands that level of speed — and chasing sub-second performance everywhere can actually be counterproductive.

It’s helpful to think about real-time in classes of latency. At the extreme, you have sub-second responses — critical for areas like algorithmic trading, where even a few milliseconds can cost millions. Then there are scenarios where a response in a few seconds is perfectly adequate — for example, fraud detection. If you can flag a suspicious credit card transaction within two or three seconds, that’s usually enough to stop the transaction before it clears. And then there are other cases where even minutes are fine. Think about a customer experience dashboard in a call center — if the screen refreshes every minute or two, that still feels live enough to be useful.

2026-01-13

This is a post in the SQL Interview series. These aren’t trick or gotcha questions, they’re just questions designed to scope out a candidate’s knowledge around SQL Server and Azure SQL Database.

Section: Administration Level: Intro

Question:

You have a database configured for simple recovery model. You notice that the transaction log is still increasing in size.

What could cause this?

Answer:

Even in the Simple recovery model, it is quite normal to see the transaction log file grow. That does not necessarily mean it’s stuck or misconfigured.

2026-01-12

When we talk about real-time, it’s worth asking: why does it matter so much? The answer is that in many situations, timing is everything. Traditional business reports are incredibly valuable, but they often arrive hours or days after the fact.

By the time you’ve spotted the issue, the opportunity to act has already passed.

Take fraud detection as an example. If a customer’s credit card is being misused, you can’t afford to wait until the next daily reconciliation. By then, dozens of fraudulent purchases might have gone through. Real-time intelligence means the system can detect unusual patterns on the spot and stop the fraud before it escalates.

2026-01-11

This is a post in the SQL Interview series. These aren’t trick or gotcha questions, they’re just questions designed to scope out a candidate’s knowledge around SQL Server and Azure SQL Database.

Section: Administration Level: Medium

Question:

You are reviewing the memory usage of a Windows Server running SQL Server. It appears to have almost run out of memory.

You add additional memory, and monitor the usage of the memory by SQL Server. It appears to always be constantly increasing, until the server is almost out of memory again.

2026-01-10