Friday, October 28, 2011

PASS Summit 2001 In Review…WOW!!!

It has been a couple of weeks and I have had time to get back in the groove at a new job as Senior SQL Server DBA, after trying Informatica Administrator via Enterprise Architect for about 3 months. Long story, but I am glad to be back as a DBA.

My week started on a Sunday travel day where I got to Seattle about 8pm Pacific and close to 10pm getting to bed at the 6th Avenue Inn after taking the Light Rail. The rail was cheap ($2.75), but took at little more work and time. The hotel suggestion was from a blog by Brent Ozar and cost ~$600 for Sunday thru Saturday.

Monday was off to Rooooobbbb Farley’s Query Tuning Pre-con. I learned a lot about Residual Predicates and GROUP BY versus HAVING versus WHERE clauses. He was ad-hoc writing queries to show performance metric and I hope to get a copy of the sql before long to review.

The second day was about networking. I started in the Virtual Chapter leaders meeting with Geoff and Karla, They did a good job informing us about options and sharing from different chapter leaders. Our Data Architecture VC can give away a $25 gift that I did not know about. If we get a sponsor, up to $100 a session. It was nice to talk with the DBA VC which had 3 people on attendance, and I got a nice shirt, that I wore while manning the VC tables during the Welcome Reception.


Tuesday was not over. Next, SQLSaturday leader meeting. This was led by Andy Warren who is a well-seasoned leader and moderator. Various suggestions were given for the website, which I think is invaluable, no matter what does not work perfectly. Anything free is nice in my opinion. Lots of information in starting a Pre-con for a SQL Saturday and talk about speaker cancelations. This seems to be a growing concern from leaders and ways were discussed to help grow more speakers and get speakers to notify about needs to cancel.


Above are pictures of Andy, Karla and the PASS IT dude. And right is Greg Larson and Sri.

The second half of Tuesday was spent at Adam Machanic’s Pre-con and performance tuning. Man, this dude knows his stuff. He and Alan White helped me understand the difference between CXPACKET problems, versus request Waits through twitter and blogs. sp_WhoIsActive is a tool you have to learn, plus read his 30 days of explanation on what it does. Adam is well organized and well rehearsed. His training is a must.

Tuesday evening was the Welcome Reception and SQLServerCentral.Com and RedGate’s Casino night. The exceptional DBA ward is given at the later event.

Wednesday morning was time to tackle the regular sessions. I made it a BI day. Started with Craig Utley from SolidQ doing SSAS Aggregates. Then, Devin Knight from PragmaticWorks presenting common SSAS mistakes. Both sessions were great and well prepared. Lunch was Chapter day and wear your SQL Saturday shirts. Jen Stirrup presented on PDW and reporting which got me wanting to investigate more into the monitoring of instances. Last was Many to many relationships in DAX by Alberto Ferrari. The last session kind of got me thinking about how to model the data to not have to write expressions in Excel. Wednesday night was Vendor appreciation dinner at the Summit. I visited with SQLSentry, Confio and Idera, ending with Quest while I had questions about Spotlight used at my current employer.

Thanks to DELL for my favorite place at PASS!!! And thanks to Robert from BCBS helping me find this treat.

Thursday morning I got to blog from the bloggers table at the morning Keynote. See previous blog for my take - It was really cool to sit next to Kevin Kline and Stacia Misner among others.

Thursday morning was Reporting Service Tips and Tricks from Bob Meyers. I really enjoyed this session. Though my favorite session on the whole week was Ami Levin talking about Physical Join operators. After I go through his examples again, and apply the lessons to some real world queries, I will have to blog about it. Do not miss this guy, he was well organized and very knowledgeable. I might be wrong about the best session, because next was Kalen Delaney talking Plan Cache. 500-600 people attending can tell you she knows her stuff. Check out her internals book for more info - Last was a chance to see a Kimball employee Joy Mundy talked about Multivalued Dimension relationships – no demos and all talk left me wanting more. Maybe my brain was already toast.

I took Thursday might off to head to my room and grab a bite. There was a need to rehearse my 3rd Normal Talk before Friday. Well rested, Friday morning started. First up, Adam Machanic again with a 500 level Query Memory session. Another great, well prepared session with demos. After lunch, I briefly went to Peter Meyers talk about KPIs in the MS BI Stack. Peter came to SQLSaturday in Baton Rouge, so I want to say hi and thank him again. All the SolidQ presenters are great in the information they share and well prepared. My talk happened fast. Only 2-3 people left out of ~150 and 5-6 questions at the end. Hope to learn from comments. I think next time, I will not go to 400 or 500 level sessions before my 200 level. I felt I did not present enough advanced stuff, but I know the database normalization talk is needed in the development community. Last, I took it easy in a T-SQL session with Aubrey Hammonds. She showed some CTEs and other cool SQL. It was a good break from the other advanced sessions. The data Architecture VC will have her presenting in the future.

The PASS Summit is a not miss event. I would rather go to it than any TechEd event. No other conference have I been able to see networking, fellowshipping and free sharing of knowledge ever. I am so glad I have been able to meet people like Grant Fritchey, Sean and Jen McCown, Andy Warren and Steve Jones just to name a few. And thank the companies that sent or let me go.

God Bless and GEAUX Tigers!!!


Thursday, October 20, 2011

Jeremy Huppatz presents Row Versioned Data Warehouse

Subject: Row Versioned Data Warehouses

Level: 200-300 (Intermediate)

Start Time: Thursday, December 15th, 2011 8:00 PM US Central Time

Presenter: Jeremy Huppatz (Blog|@OzzieMedes)

Live Meeting Link:

Row Versioned Data Warehouses
Jeremy will discuss the strengths and disadvantages of a row-versioned data warehouse design in the context of a real world case study, sharing lessons learnt and demonstrating some of the technologies and techniques used to build a row-versioned data warehouse. He will discuss row-versioning in the context of Kimball slowly-changing dimensions, and will also provide some details on the relative strengths of row-versioning as applied to measure information in situations where such data cannot be considered finalized at the time it is loaded into the warehouse.

Jeremy Huppatz
Jeremy has been a SQL Server data guy going back as far as 1997. During his 14 years working with Microsoft’s flagship database, he has filled the roles of DBA, data modeller, database designer, data analyst, BI developer and data architect in a multitude of projects, which have included departmental database apps, enterprise data warehouses and a bit of just about everything in between. Jeremy now runs his own IT consulting firm called Solitaire Systems and lives in the scenic Adelaide Hills with his partner Alison and her cat. He also plays, writes and records music and writes in his spare time, and has been described optimistically as an avid (as opposed to obsessive) computer gamer.

The PASS Data Architecture Virtual Chapter will focus on data architecture concerns as they impact users, developers and DBAs on the Microsoft SQL Server platform. We want to make data architecture accessible to all data practitioners, and drive the point home that Data Architecture is a set of practices and a body of knowledge that overlaps almost all database professionals to some degree.

Thursday, October 13, 2011

SQL PASS 2011: Live Blog Keynote Day 2–Quentin Clark

7:50AM – arrived at bloggers table and positioned between Rooob Farley and Kevin Kline. Just meet SQLBalls, in front is Ryan Adams and Mike Walsh.

Hopefully, this will be a demo Keynote and not just talk, talk, talk…

8:10AM – Room is filling up, cannot believe so many people here at #SQLPASS, they are saying a 22% increase from last year. All the rooms are pack, some have people on the floor. Andy Warren and Stacia just sat down and are typing and talking. John Sterret just came and said hi.

8:25AM – Laptop crashes OOOHHH!!!!

8:30-8:40am – Bill talks about the wonderful volunteers around the country for SQL Server. Lori Edwards receives PASSion award. Tim and Jeff mentioned about their work

8:45AM – Quentin Clark invited Bob Erickson from Interlink Transport – talks about operations and uses of SQL Server (Mission critical is Their Business)

8:50AM – demonstrating Always ON for Interlink – Wizard setup and ASync versus Sync. Listening is a DNS name and connect to name and does the redirecting

8:55AM – on to Blazing fast speeds – VertiPak built into Analysis Services, takes advantage of the technology and in the RDBMS. Some customers are reporting 10, 20 & 30 times processing speed.

9:00AM – Power View + PowerPivot – SSAS on the back-end but control by IT (SharePoint). Alerts on reporting, configurable. A BI Semantic Model – ‘who do you know…’ Semantic understanding. As the model is used, the engine learns from queries. Now, to get it in a data model. Master Data Services – managing the location of where everything is located.

9:05AM – @SQLGal is on stage. Using Contoso database. Columnstore index demonstration on query in report going from 45 seconds to .13 seconds after ColumnStore index and linking data to Cloud location for objects in map not appearing where there are actually located. Also,

9:10Am – Parallel Data Warehouse – Appliances. Deep Dive to workloads, and understand characteristics and find the architecture to solve problems. Appliance developer kit to fit customer. Demos DELL parallel warehouse, more than one appliance.The stage has the DELL racks for PDW. And th HP rack that has been out over a year. Also, have your own Private Cloud.

9:20AM – ODBC drivers to Linux – more and better drivers.

9:25AM – integrating CDC with Oracle, SQL Server 2012 has it – Now Beyond relational. file Stream 2D Spatial, Semantic search

Beyond full-text search – better to be seen – Demo time

The demos need some Zooming

Semantic search extracts the keywords from the documents. More than Full-text search.

9:30pm – ‘Juneau’ – now SQL Server data Tools -

Now embed SQL Server Express DB in an Application

Scaling to the Cloud – spin up many many instances in the cloud to distribute queries – DEMO time

Nicholas Dritsas SQLCAT team – SQL Azure demo

9:35AM – Who is @CajunSQL – did not know of any other CoonAxxes here at PASS

9:40AM  - Real customer demo with SQLAzure – moved to cloud form

9:45am – Using Federation in the Cloud – sizing and scaling in the cloud

Example is a blog site that does not know what blog goes viral – Blogs r’ Us

No developer rework of code

Neat interface/management tools – Schema walker

Overview of tasks going on – Powerful reporting

Ability to create large database – up to 150 GBs in size

9:50am – Azure and SQL in same code base.


God Bless,

Thomas Smile

Monday, October 10, 2011

PASS Summit 2011: Day One

After getting a little lost from the rail to the hotel, I finally got to sleep Sunday night after a 10+ hour airplane/airport travel time. Well worth it especially after seeing LSU whoop up on Florida in Tiger Stadium on Saturday afternoon.

Started by meeting Robert from BCBS to have an early registration and breakfast and the networking started at 7:30am with a gentleman from New York and another from California. We chatted about performance tuning on SQL Server as well as Sybase. Rob started to see what I meant about the value of PASS right away. He went to Brent Ozar’s SANs and Virtualization Pre-con and I hit Rob Farley’s 1/2 comedy/music show and Advanced TSQL for query tuning.


Lunch break had Robert and I at a table with Ryan and David from Dallas talking about SQL Saturday’s. We also had some new friends for lunch to chat about different companies we worked for.

The second half of the day was more Rob Farley and TSQL. He was able to explain some property window information about query plans. Never realized sometimes you have to read the plan from left to right to really understand them. WHERE and HAVING clauses were explained in more detail. The highlights was a new term for me: Residual Predicates. You really need to read up on this to understand advanced query tuning.

Rob even helped promote my 3rd Normal Talk on Friday.

The evening was a well organized Networking dinner that Andy Warren and Steve Jones put together at Lowell’s. The food line moved at a good pace for outside conversations, but the food was served fast and it was good. Saw @SqlDiva from New York and Kathy K from Microsoft. Idera was there with Dave and Heather.

Robert and I called it a night at 10PM CST (we are on Pacific time, so it was 8pm), but we were both tired.

Tomorrow is all about networking with Virtual Chapter leaders, local user group leaders and SQL Saturday organizers. Should be interesting.

Here are some links:

Rod Farley Residual Predicate -

Lowell’s in Seattle -

24-Hours of PASS DB SOP - Standard Operating Procedure for Normalized Database Design

God Bless,

Thomas LeBlanc Smile