SQL Relay 2014 Bristol

This was the third 2014 event in the SQL Relay.  SQl Relay Bristol was held on 15 October at Future Inns, Bristol.

The Microsoft keynote covered the question of where next after big data. There is a driving force now focusing on data and data culture, looking at how we can leverage an edge and be predictive. How can diverse data be combined together with analytics and people at speed. The three areas  

·         Age of the Algorithm – Machine learning

·         Age of Things

·         Age of Intelligence   – Intelligent systems with ubiquitous computing

Microsoft shared details of a Data Insight Immersions event taking place around the country, details of which can be found here - http://bit.ly/1waXnuD

Some useful resources from the event are listed here:

SQL Server 2014 In-Memory

Some good resources for learning about SQL Server 2014 In-Memory

Channel9 Data-Exposed

http://channel9.msdn.com/Shows/Data-Exposed

In-Memory OLTP – Common Workload Patterns and Migration Considerations whitepaper http://technet.microsoft.com/en-us/library/dn673538.aspx

Microsoft DocumentDB

Azure DocumentDB is a fully managed service with fast reads and writes and flexible schema. It natively supports JSON and has a simple API for interacting with the data. An introduction can be found

http://azure.microsoft.com/en-gb/documentation/articles/documentdb-introduction/

 

Azure Machine Learning

This session on machine learning covered the resurgence in predictive analytics highlighting the benefits for making a breakthrough in machine learning.  The conceptual architecture, data science process and data mining models used were introduced.  The main thing to take away is that BI is reactive whereas machine learning is predictive. The design environment is free to use when you are not processing data. Training available free can be found

Microsoft - http://azure.microsoft.com/en-gb/services/machine-learning/

Machine Learning course at www.coursera.org

The Johns Hopkins Data Science Specialization http://jhudatascience.org/

Other great session included SQL partitioning design, SQL Azure Backup to the Cloud etc.

Database Readling Lists

I keep coming across reading lists so I thought I would start colating these for ease of reference.

Collated Database Reading list 

This reading list is quite comprehensive covering

• Basics and Algorithms
• Essentials of Relational Databases
• Classic System Design
• Columnar Databases
• Data-Parallel Computation
• Consensus and Consistency
• Trends (Cloud Computing, Warehouse-scale Computing, New Hardware)
• Miscellaneous
• External Reading Lists

https://github.com/rxin/db-readings

Big Data and Enterprise

The U.S. Chamber of Commerce Foundation has compiled a reading list related to Big Data and data-driven innovation.

http://www.uschamberfoundation.org/reading-list-big-data-and-enterprise

Best Machine Learning Resources for Getting Started

http://machinelearningmastery.com/best-machine-learning-resources-for-getting-started/

Data Lakes

The term Data Lakes keeps appearing in conjunction with new data platforms. I came across this article from Gartner who says beware of the data lake fallacy. They define data lakes as:-

"In broad terms, data lakes are marketed as enterprisewide data management platforms for analyzing disparate sources of data in its native format," said Nick Heudecker, research director at Gartner. "The idea is simple: instead of placing data in a purpose-built data store, you move it into a data lake in its original format. This eliminates the upfront costs of data ingestion, like transformation. Once data is placed into the lake, it's available for analysis by everyone in the organization." http://gtnr.it/1E6d73V

SQLBits XII

SQLBits XII held at The International Centre, Telford between 17-19 July was an event that made history and will be the talk of the town for years to come. A big thank you needs to go the committee, speakers, delegates, sponsors and helpers.This is the forth year I have helped a SQLBits and it is always such a privilege to help at the event.

The session content stretched the breath and depth of the SQL Server platform from Big Data Analytics, Business Intelligence, Development and Administration. 

I spent the training day at the Real World Big Data-Deploy, Design and Manage like a Pro session with Adam Jorgensen and John Welch. The session linked Hadoop to BI solutions providing examples in labs culminating in streaming real time analytics. It was packed full of tips on best practices, configuration patterns, predictive analytics for the analysts and BI integration for the data developers. These new technologies allow you to modernize your data architecture. This was a great practical session. 

The keynote was delivered by Nigel Ellis a Distinguished Engineer in the Database Systems Group at Microsoft. It continued along the track of previous technical keynotes starting with an overview of Azure SQL Database progressing to the technicalities of ring geometry of distributed computing.
 
Greatly enjoyed the session on Latches, Spinlocks, and Lock Free Data Structures. It is always fun to delve into this area using sys.dm_os_latch_stats which returns information about all latch waits organized by class.

Microsoft gave a session on In-memory tables which introduced  the in-memory technology for optimizing OLTP workloads and gave some key pointers to consider when designing new architectures. There are currently some scope limitations. The document Common Workload Patterns and Migration Considerations is a good source of reference.

There were many, many more sessions delivered by excellent speakers and it is never possible to attend them all.  It is good to know the sessions are recorded and will be avaialble later on the SQLBits website.

A few mentioned a useful paper to read by Joseph Sack on the new SQL Server 2014 Cardinality Estimator.The article is entitled Optimizing Your Query Plans with the SQL Server 2014 Cardinality Estimator . This is an important new feature that can make existing plans perform the same, better or worse so it is key to test these changes out. You can use trace flag 2312 and 9481 to test code using the different optimisers. The article Enable plan-affecting SQL Server query optimizer behaviour that can be controlled by different trace flags on a specific-query level describes the flags.
 

  

Pre-SQLBits - Bristol SQL Server User Group - July Meeting

Monday 14 July saw the Bristol User Group meet with Brent Ozar talking on Are AlwaysOn Availability Groups Right for You? It was amazing that Brent came to share his in depth knowledge on AlwaysOn. He shared the pros and cons of what needs to be considered when creating this type of database architecture solution.

Brent has an amazing set of resources covering AlwaysOn. These are  found
http://www.brentozar.com/sql/sql-server-alwayson-availability-groups/ 

The Microsoft guides I found useful were:-

AlwaysOn Architecture Guide: Building a High Availability and Disaster Recovery Solution by Using AlwaysOn Availability Groups
http://msdn.microsoft.com/en-us/library/jj191711.aspx

Microsoft SQL Server AlwaysOn Solutions Guide for High Availability and Disaster Recovery
http://msdn.microsoft.com/en-us/library/ff929171.aspx

Comparison of Failover Cluster Instances and Availability Groups
The core concepts between FCIs and Availability Groups are compared in this table taken from http://msdn.microsoft.com/en-us/library/ff929171.aspx
 

For more details and caveats see the full article http://msdn.microsoft.com/en-us/library/ff929171.aspx

Mainstream Support Ends

Mainstream Support for SQL Server 2008 and SQL Server 2008 R2 will end on  8 July 2014.  The definition of Mainstream support is shown in this table. 


Taken from http://support.microsoft.com/lifecycle