SQL Server Denali CTP3 and Data Mining
Today, Microsoft has released the CTP3 version of the next version of SQL Server, codenamed Denali. The acronym CTP stands for Community Technology Preview. The purpose of this preview is to encourage people outside Microsoft to use the software before Microsoft releases an official version.
I, along with some select others, have been participating in the early access program called “Technology Adoption Program” (abbreviated TAP). Participation in Microsoft TAP is by invitation only, and had included CTP1 and CTP2. (Now that Microsoft invited me into the Microsoft MVP program, I am likely to be invited to similar opportunities in the future.) However, today’s release is widely available for EVERYONE (that means you too).
The purpose of this blog post is to describe the location of the new download, and encourage you to try it out for SQL Server Data Mining.
Microsoft provides some information about Denali at http://www.microsoft.com/sqlserver/en/us/product-info/future-editions.aspx
Please download the CTP3 from: https://www.microsoft.com/betaexperience/pd/SQLDCTP3CTA/enus/default.aspx
In addition to the core software, there are separate downloads which comprise the “Feature Pack”. These other pieces are not included in the main download, but have additional functionality which not everyone requires. All these other individual Feature Pack elements are licensed by Microsoft, and some free but others would require SQL Server to be fully functional.
(Download Feature Pack for CTP3 is no longer available)
Microsoft® SQL Server® code name ‘Denali’ Community Technology Preview 3 (CTP 3) Feature Pack — List of Features
- Microsoft® SQL Server® code name ‘Denali’ Master Data Service Add-in for Microsoft® Excel® CTP 3
- Microsoft® SQL Server® code name ‘Denali’ Semantic Language Statistics CTP 3
- Microsoft® SQL Server® Report Builder for Microsoft® SQL Server® code name ‘Denali’ CTP 3
- Microsoft® SQL Server® code name ‘Denali’ PowerPivot for Microsoft® Excel CTP 3 Microsoft® SQL Server® code name
- ‘Denali’ Reporting Services Add-in for Microsoft® SharePoint® Technologies
- Microsoft® SQL Server® code name ‘Denali’ Reporting Services Add-in for Microsoft® SharePoint® Technologies
- Microsoft® SQL Server® code name ‘Denali’ Data-Tier Application Framework CTP 3
- Microsoft® SQL Server® code name ‘Denali’ Transact-SQL Language Service CTP 3
- Microsoft® SQL Server® code name ‘Denali’ Transact-SQL ScriptDom CTP 3
- Microsoft® SQL Server® code name ‘Denali’ Transact-SQL Compiler Service CTP 3
- Microsoft® SQL Server® Compact 4.0
- Microsoft® SQL Server® Compact 4.0 Books On-line
- Microsoft® SQL Server® JDBC Driver 4.0 Community Technology 2 (CTP 2)
- Microsoft® Connector 1.1 for SAP BW for SQL Server® code name ‘Denali’ CTP 3
- Microsoft® System CLR Types for SQL Server® code name ‘Denali’ CTP 3
- Microsoft® SQL Server® code name ‘Denali’ Remote Blob Store CTP 3
- Microsoft® SQL Server® code name ‘Denali’ Books On-line CTP 3
- Microsoft® SQL Server® code name ‘Denali’ Upgrade Advisor CTP 3
- Microsoft® SQL Server® code name ‘Denali’ Native Client CTP 3
- Microsoft® OLEDB Provider for DB2 v4.0 for SQL Server code name ‘Denali’ CTP 3
- Microsoft® SQL Server® code name ‘Denali’ Command Line Utilities CTP 3
- Microsoft® SQL Server® Service Broker External Activator for SQL Server® code name ‘Denali’ CTP 3
- Microsoft® Windows PowerShell Extensions for SQL Server® code name ‘Denali’ CTP 3
- Microsoft® SQL Server® code name ‘Denali’ Shared Management Objects CTP 3
- Microsoft® SQL Server® code name ‘Denali’ ADOMD.NET CTP 3
- Microsoft® Analysis Services OLE DB Provider for Microsoft® SQL Server® code name ‘Denali’ CTP 3
- Microsoft® SQL Server® code name ‘Denali’ Analysis Management Objects CTP 3
- Microsoft® SQL Server® Driver for PHP 2.0
- Microsoft® SQL Server® Migration Assistant (which has five parts)
1. Microsoft SQL Server Migration Assistant for Access
2. Microsoft SQL Server Migration Assistant for MySQL
3. Microsoft SQL Server Migration Assistant for Oracle
4. Microsoft SQL Server Migration Assistant for Sybase
5. Microsoft SQL Server Migration Assistant 2008 for Sybase PowerBuilder Applications
I recommend installing this software on Windows Server 2008 R2 with Hyper-V virtualization. You could also use another virtual environment (many people use VMWare). If you do not want to use Windows Server, then my next choice is Windows 7. Alternatively, you could reimage a new machine just for this installation.
I would NOT recommend running this version on a machine which has another version of SQL Server. An issue I had is that double installation of Denali CTP1 on the same computer as SQL Server 2008 R2 messed up Business Intelligence Development Studio (BIDS). However, someone who earned the Microsoft Certified Master for SQL Server (hint: he is from Australia) says that Microsoft would like to hear from people doing multiple version installations. And yes if you have the depth of knowledge about SQL Server, and are willing to fix anything that might break, then yes, go for a multiple installation.
I wondered if Microsoft recommends multiple version SQL Server installations for production servers. Here’s what my Master friend says:
As a heads-up, they do support multiple versions in production environments. It’s actually quite common and is often how migrations to new versions happen.
The biggest issue is around the client tools. Only major versions are usually supported side-by-side in client tools. For example, you can install SQL Server 2008 R2 client tools beside 2005 client tools. But you can’t install SQL Server 2008 R2 client tools beside 2008 client tools as they have the same major version. 2008 R2 would just upgrade your 2008 tools if you tried to do that.
SQL Server 2012 tools *should* work side-by-side with either 2005 or 2008/R2 but how well this works remains to be seen
Thus, I DO recommend trying this version in production environments. Note that this recommendation is an opinion, and should NOT be considered advice. Install at your own risk, though people who are technically experienced know how to manage this type of risk. I have not earned the Microsoft Certified Master, but could point you to some people who have.
Your Feedback on Connect.Microsoft.Com
If you have an issue, or any feedback from the functionality or usability of the interfaces, then Microsoft wants to hear from you. I recently heard from Doug Turnure, who now manages the Microsoft Connect website, which is for all bugs and suggestions, not just SQL Server. His team has streamlined their process of feedback for all products (as recently reported to the audience attending Atlanta Code Camp). The link for feedback specifically for SQL Server (including Denali): https://connect.microsoft.com/sqlserver/feedback
To my knowledge, there are no new data mining features in this Denali version compared with SQL Server 2008 R2. While I do not know of any issues with data mining, and while I expect the data mining technology to work reliably, it is remotely possible that something might not work (and if you discover something, please post it to the Connect website).
This version has many other new data management features. As I have been saying about SQL Server Data Mining, you purchase not just a service targeted for enterprise production use, but also are able to leverage any of the other data management features which come with SQL Server. I recommend looking at the new features in Denali.
Why Enterprise Data Mining Requires Enterprise Data Management
Let me step back to the larger picture. Experienced data miners recognize that most of their time is spent in the ETL phase (Extract-Transform-Load), which typically includes data cleaning too. “Data mining” describes activity for the real world, while “machine learning” often focuses specifically on algorithm efficiency. In working with clients, I take a larger approach, and discuss issues all the way from data acquisition to final information presentation. My mentor W. Edwards Deming advises people to take a view of organizations that recognizes the entire system, and not just the technology. Systems thinking is important for successful data mining because the performance gains by either a machine learning algorithm or an ETL process could be completely removed or reversed by other organizational factors.
Just focusing on technology, production data mining solutions require improvements in data management and access. This new release of Denali has many enterprise enhancements which can help improve the entire data mining process. I have only selectively blogged about some publicly-announced features such as the Apollo Columnstore Indexes (which as it turns out has been one of the most often read posts on this blog). However, with this public CTP3 release, I can now freely engage in discussion about whatever is in the released Denali CTP3.
If there are any specific features or questions you would like me to answer or discuss, please use the contact form and send me a message.