Press "Enter" to skip to content

Day: February 17, 2020

An Overview of Generative Adversarial Networks

Mohammad Waseem takes us through an overview of Generative Adversarial Networks:

Generative models are nothing but those models that use an Unsupervised Learning approach. In a generative model, there are samples in the data i.e input variables X, but it lacks the output variable Y. We use only the input variables to train the generative model and it recognizes patterns from the input variables to generate an output that is unknown and based on the training data only.

In Supervised Learning, we are more aligned towards creating predictive models from the input variables, this type of modeling is known as discriminative modeling. In a classification problem, the model has to discriminate as to which class the example belongs to. On the other hand, unsupervised models are used to create or generate new examples in the input distribution.

To define generative models in layman’s terms we can say, generative models, are able to generate new examples from the sample that are not only similar to other examples but are indistinguishable as well.

Click through for the overview.

Comments closed

Benchmarking Apache Hadoop Ozone

Istvan Fajth and Mukul Kumar Singh take us through a benchmarking test of Apache Hadoop Ozone:

Apache Hadoop Ozone was designed to address the scale limitation of HDFS with respect to small files and the total number of file system objects. On current data center hardware, HDFS has a limit of about 350 million files and 700 million file system objects. Ozone’s architecture addresses these limitations[4]. This article compares the performance of Ozone with HDFS, the de-facto big data file system. 

We chose a widely used benchmark, TPC-DS, for this test and a conventional Hadoop stack consisting of Hive, Tez, YARN, and HDFS side by side with Ozone. True to the current industry need for separation of compute and storage, which enables dense storage nodes and elastic compute, we run these tests with the datanodes and node managers segregated. The fundamental ambition of this endeavor, and the subsequent effort in optimizing the product, is to be comparable in terms of stability and performance to HDFS. To that end, we would like to call out the amazing amount of work put in by the community over the past several months towards this goal.

It’s interesting to watch the Hadoop community work through these sorts of challenges, where the hardware paradigm has differed quite a bit from when HDFS was created.

Comments closed

Monitoring Power BI Data Gateway Performance

Olya Musokhranova shows how we can track throughput and utilization of the Power BI On-Premises Data Gateway:

In June 2019, Microsoft made Gateway Monitoring much easier via the addition of structured logs that can be enabled in the configuration file, and by providing a starter Gateway Performance Power BI report template to visualize the results.

These logs give full access to the information, minimizing the need to pull counter data from IT management systems, like the Microsoft System Center Operations Manager (SCOM); or run the Windows Performance Toolkit and dig through dust-covered gateway info and error log files; or pull refresh error messages from the Power BI Service. Enabling structured Gateway Monitoring logging doesn’t add any significant load to the server, so that’s another win.

Read on for a schema and inspiration on how to set up and configure your own dashboard.

Comments closed

NULL Handling Features not in T-SQL

Itzik Ben-Gan continues a series complexity around NULL:

When using the offset window functions LAG, LEAD, FIRST_VALUE and LAST_VALUE, sometimes you need to control the NULL treatment behavior. By default, these functions return the result of the requested expression in the requested position, irrespective of whether the result of the expression is an actual value or a NULL. However, sometimes you want to continue moving in the relevant direction, (backward for LAG and LAST_VALUE, forward for LEAD and FIRST_VALUE), and return the first non-NULL value if present, and NULL otherwise. The standard gives you control over this behavior using a NULL treatment clause with the following syntax:

offset_function(<expression>) IGNORE_NULLS | RESPECT NULLS OVER(<window specification>)

There are three good examples of functionality around handling NULL which the current implementation of T-SQL is missing.

Comments closed

T-SQL Checker ADS Extension

Daniel Janik has a new Azure Data Studio extension for us:

I’ve created a sample ads extension that checks TSQL syntax in real-time for potential bad practice.

Right now the extension is using regex, which isn’t the best for parsing SQL but it works for the few test cases I’ve added.

I’m hoping that the community can help evolve the project into something really cool; so, I’m asking for your help in making that happen.

Click through for a demo, and check out the GitHub repo.

Comments closed

Power BI Query Diagnostics

Paul Turley has a video covering Power BI query performance:

This post demonstrates how the order of steps added to a query can make a big performance difference and drastically effect the number of steps generated by the designer. I’ll demonstrate how to use the new query Diagnostics tools to compare and understand query performance.

The Power Query Editor for Power BI simplifies data transformation processing by generating query steps for each action you perform in the query designer. This whiteboard diagram shows the high-level flow of information through a Power BI solution. Every query has a source (“SRC” in the diagram) followed by a connection. The query consists of a series of transformations (“XForm”) prior to populating a table in the data model.

Read on for a high-level explanation followed by a video which covers the Query Diagnostics feature.

Comments closed

Storing Power BI Audit Logs in Blob Storage

Gilbert Quevauvilliers works around a built-in constraint with Power BI Audit Logs:

With the new Power BI Get-PowerBIActivityEvent I wanted to find a way where I could automate the entire process where it all runs in the cloud.

One of the current challenges with the Audit logs is that they only store 90 days, so if you want to do analysis for longer than 90 days the log files have to be stored somewhere. Why not use Azure Blob Storage?

Whilst these steps might appear to be rather technical if you follow them and you have access to an Azure Subscription you can do this too.

Gilbert warns us up-front that this will be a lengthy post and that is quite true. But if you need to hold those audit logs more than 90 days, this is a great way of doing so.

Comments closed

Explaining the HA/DR Licensing Changes

Kevin Chant goes into the fairly recent licensing changes for SQL Server:

Which surprised me a bit because these licensing changes have been in-place for a while now. With this in mind, I thought I would discuss them here to raise awareness about the changes.

To clarify, in SQL Server 2019 there have been some big licensing changes about what you can and can’t do on a passive fail-over instance. Especially if you have Software Assurance.

Which I have to admit I am really excited about. Because it opens up some new possibilities which I will explain below. Of course, there are other significant updates in the licensing guide as well.

Read on for the details.

Comments closed