I have been on the “What’s New In CTP1 for Denali” speaking “tour” (if you count 2 as a tour) for the past week. What I can say from my interactions, and the feedback from the presentation, there is some real excitement for the new release (although dauntingly far away for some people). In my presentation I go over the “what’s new” for SSIS… and finish up with a look at the new dependency services piece. It is clearly the dependency services that has the BI folks hooked! To support the interest that people have expressed, I wanted to write a quick post to cover some of the dependency services parts, and get you rolling with what will be a very cool, very important part of the BI infrastructure in Denali.

Install

The installation is not complicated, however, is not available from within SSMS. The command line documentation from MSDN is easy to follow, and will get you running with a few simple commands.

Once you have it installed, you need to create extraction points for the things you want to look at. There is a wizard (right-click) that will get you started with the configuration.

In my example, I created a “test” extraction point for SSIS that is pulling from a project that I have already deployed to the server. By right-clicking on that extraction point, you now have two steps to take:

First, you need to select “update catalogs”, then right-click again and select “Sync Now”. Once you have done these two steps, double- click the “default” under views, and you will be able to navigate through the dependencies that you have defined in your extraction points.

Conclusion

I believe you will very quickly feel comfortable navigating the structures, viewing dependencies, and gaining an appreciation for how quick you will be able to do impact analysis with this new tool.

 I have to say that this has not turned into the “series” that I projected so optimistically as I departed for the PASS summit. It may still be a series, who knows, but for now, as I sit in the airport waiting for my flight back home I figure I would get a few thoughts together on the sessions that I attended.

 I tried to hit a cross section of things I understand, and would like to do better, things I really need to learn, and a few things that I am not sure how they apply to my life (education for the sake of it). I think I hit that goal…

Tuesday

AD371S – Grant Fritchey (one of the smarter people you will run into) presented “Identifying and Fixing performance problems using execution plans.” This one fits into the area of things I know, and just wanted to get a smart persons perspective, and validate some of the things I do. An excellent sessions, filled with great discussion well done, and well worth the time!

BID275M – Pej Javaheri, Lynn Langit, Donald Farmer (actually had a stand in for Donald… unfortunately, don’t remember his name right now), and a few others… presented “Business Intelligence Power Hour” – (should be named the comedy hour) This was fun (and educational) Short witty presentations on the different tools in the BI space, what is new in Denali, and why BI can help solve problems like why people with higher taxes that drink more are happier. This fit a few of my criteria of things I know, and things I would like to know better… and besides it was fun!

DBA237 Aaron Nelson PACKED the room with over 380 people for “The Dirty Dozen: PowerShell scripts for the Busy DBA” This one is in the area of things I need to learn how to do, and Aaron presented some scripts, and REALLY made the whole thing easy to understand. I honestly think we scared him a bit, this was his first PASS presentation, and a (literally) standing room only crowd was a bit intimidating, but he hung in there, and did really well with it.

DBA391S Kevin Kline presented “End-to-End Troubleshooting for SQL server” Kevin is always good to listen to, and this was no exception. My head was rather full by this time, so I decided to go with another topic that I am comfortable with, and just validate some of my methods and assumptions about how things are.

Wednesday

DBA388S Grant Fritchey presented “DMV’s as a shortcut to Procedure tuning” Since Grant’s presentation on Tuesday was so good, I figured I would jump back into learning mode with the DMV’s. Something I need to do better, and Grant is just the guy to get you there.

 AD311 Rob Farley had an absolutely OUT OF CONTROL funny educational BLAST of a sessions “The Incredible Shrinking Execution Plan” I really did no know what I was in for with this one… but Rob had us rolling the whole time! He literally broke about every “rule” of presenting (including the never type in a presentation)… he started off with a blank SSMS window, no slides, and a comedy routine that was some of the funnier stuff of the conference. With all that said, by the end of the presentation, my head HURT with the amount of things I need to evaluate how I look at queries, views, and how the optimizer views joins in general.

BIA380M Matt Masson presented “What’s Coming Next in SSIS” This one fit simply because SSIS is a core part of what I do, and Denali is changing, fixing, enhancing etc a huge portion of how SSIS works. After this session, I was truly (really) upset that we are only on ctp-1, and it will be sometime next year before we can actually push what is a drastic change to production.

PD163 Christine Valdes presented (with some help from Brent and others) “SQL Image Wardrobe Governor: The Newest Feature in R2” This one fit in the… I can’t think anymore, Brent is funny, and why not do something for the sake of education. Fun interactive session talking about why you dress the way you do (or should).

Thursday

Keynote address from David Dewitt. I have to say he is likely the smartest person that I have ever been in the same room with. I suddenly have more faith in the query optimizer than just about any technology we use on a daily basis. He really tried to talk to us in a way we could understand… and left us ALL in the dust so many times it was crazy. I wish we could have more presentations from folks like him, even if I don’t totally understand, something might just rub-off and stick in my pea-brain. Truly an impressive keynote, that I will be watching (a few times) on the DVD set.

BIA379S Marco Russo presented “Monitoring Cube Performance and Usage”

This starts getting into the area of things I need to learn. Analysis services, performance of cubes etc. A good presentation, a percentage of which I need to re-watch when I get the DVD’s. This session came right after the keynote address for David DeWitt, my mind was a bit liquified when we started this session.

BIA206 Stacia Misner presented “Real World Analysis Services Stored Procedures” This was jumping neck deep into things that I don’t understand enough to be able to use (at this level). This is one was mostly education for the sake of education. Great presentation, MDX, and stored procedures in analysis services, DEEP stuff.

DBA247 Ken Simmons presented “Enforcing Compliance with Policy-Based Management”

This one I went to for a few reasons. I use policy based management, and wanted to see Ken’s take. I needed something to make me feel like I actually know something (the keynote, followed by cube performance, followed by SSAS procedures… I needed something to boost me back up) and not the least of the reasons, is I really like Ken, and he was stressed about the session not having enough questions, so I came to support him (and only had to ask one question, there were PLENTY from the audience) and Ken handled the session like the pro that he is.

BID216 Andreas Wolter presented “Report Builder 3 What’s in it For You” At this point in the conference I have to say my tank was empty. Late nights (that might be another post… maybe), early mornings, and more information crammed into my head than I have had in a LONG time. So I figured I would wrap up the day with a session that I understand, pick up a few pointers (which I did), and think about how we might get some of the spacial data included in the work that we currently do.

Conclusion

I can honestly say this was a great (understatement of the week) conference. I learned a ton, learned that I still have a ton to learn, and met some of the nicest, smartest people you will ever meet. It was great to put faces to many of those that I talk with regularly, and meet some great new friends. Not to mention the fact that I now have a bunch of work to do to get the presentations ready that I have promised to a variety of folks, including my new friends in Australia. One heck of a week…

Normally, my blog is technical in nature, and (hopefully) generally helpful. This series will likely be nothing like that…. maybe helpful for something, but most likely not. As I sit and write this introduction, I am aboard a United flight looking out the window at the Appalachian mountains en-route to Chicago and then on to Seattle for a few (expecting to be hectic) days at the SQL PASS conference.

So I guess there are a few reasons for this series… not the least of which is to have something to look back on and reflect what I learned. So sit back, put your tray tables in the upright position, fasten your seat-belts, and hopefully this will be an informative trip for all of us.

The obvious part-1 of the blog is travel…. I know there are LOTS of road worriers out there that spend (most) of their time on the road. I imagine there is little that I can actually add in the travel department, other than carry as little “stuff” through airport security as you can manage. I have a backpack for my carry-on/briefcase, which needed to be EMPTIED of the contents into little plastic bins… netbook, charger, camera, cellphone. Not to mention the removal of shoes, belt, jacket, everything from your pockets…. makes me wonder why we need the full-body scanner at all.. As for on-the road technology, I am keeping it relatively simple with a netbook. I know there are LOTS of you ipad folks out there, and maybe someday I will come to your side, but for now, something that I can use to type on, read on, remote to work on, take notes on.. still seems like a better choice. I know there are apps for just about all of that, but the lack of actual keyboard is still a problem for me. As for apps… evernote is my personal favorite choice. I have gone (largely) paperless because of evernote, there is little that I don’t store there (that is likely a good post for another day). All the conference docs (which I need to go read to figure out what sessions I am attending). All of the travel docs, and what will be all of the conference notes.

So I guess, that is it for part-1, I hope the rambling of a conference in Seattle will at least be entertaining… but if nothing else, it will give me something to reflect on, and you can come back for the more technical and informative content when I get back from the trip.

I have seen several posts lately asking about ROI/Justifications etc about upgrades. There are LOTS of posts on this topic, and some really great documentation from Microsoft that covers this topic at great length. This post does not really add anything to those, other than distilling some of the key points from the Microsoft documentation that I used when discussing the upgrade, benefits, and expectations with my business leaders.

Summary

With the 2008 release, SQL Server makes a major advance in scalability for data warehousing. It meets the data warehouse needs of the largest enterprises more easily than ever. SQL Server 2008 provides a range of integrated products that enable you to build your data warehouse, and query and analyze its data. These include the SQL Server relational database system, Analysis Services, Integration Services, and Reporting Services. This paper introduces the new performance and manageability features for data warehousing across all these components. All these features contribute to improved scalability.

Upgrade Introduction

Microsoft® SQL Server™ 2008 provides a comprehensive data warehouse platform. It enables you to build and manage your data warehouse, and deliver insight to your users, with a single, integrated product suite. It scales to meet the needs of the largest enterprises, in a way that empowers both your end users and your IT staff.

The number one focus of development in the SQL Server 2008 release was to improve scalability across the entire product suite to comfortably meet the needs of large enterprises. Here, we’ll introduce the features and enhancements we’ve added to improve your data warehouse experience. Build. Manage. Deliver. SQL Server 2008 lets you do it all, with ease.

Map of New Data Warehousing Features

The following table shows the new scalability features in SQL Server 2008, and where they help with the activities that surround your data warehouse (DW).

Data Compression

The new data compression feature in SQL Server 2008 reduces the size of tables, indexes or a subset of their partitions by storing fixed-length data types in variable length storage format and by reducing the redundant data. The space savings achieved depends on the schema and the data distribution. Based on our testing with various data warehouse databases, we have seen a reduction in the size of real user databases up to 87% (a 7 to 1 compression ratio) but more commonly you should expect a reduction in the range of 50-70% (a compression ratio between roughly 2 to 1 and 3 to 1).

Backup Compression

Backup compression helps you to save in multiple ways.

By reducing the size of your SQL backups, you save significantly on disk media for your SQL backups. While all compression results depend on the nature of the data being compressed, results of 50% are not uncommon, and greater compression is possible. This enables you to use less storage for keeping your backups online, or to keep more cycles of backups online using the same storage.

Backup compression also saves you time. Traditional SQL backups are almost entirely limited by I/O performance. By reducing the I/O load of the backup process, we actually speed up both backups and restores.

Resource Governor

The new Resource Governor in SQL Server 2008 enables you to control the amount of CPU and memory resources allocated to different parts of your relational database workload. It can be used to prevent runaway queries (that deny resources to others) and to reserve resources for important parts of your workload. SQL Server 2005 resource allocation policies treat all workloads equally, and allocate shared resources (for example, CPU bandwidth, and memory) as they are requested. This sometimes causes a disproportionate distribution of resources, which in turn results in uneven performance or unexpected slowdowns.

Star Join

With dimensionally modeled data warehouses, a big part of your workload typically consists of what are known as star join queries. These queries follow a common pattern that joins the fact table with one or several dimension tables. In addition, star join queries usually express filter conditions against the non-key columns of the dimension tables and perform an aggregation (typically SUM) on a column of the fact table (called a measure column). With SQL Server 2008, you will experience significant performance improvements for many star join queries that process a significant fraction of fact table rows.

Partitioned Table Parallelism

Wouldn’t you like to get the most power you can out of the hardware you own? The partitioned table parallelism (PTP) feature in SQL Server 2008 helps you do that. Data warehouse applications typically collect large amounts of historical data in fact tables, which are often partitioned by date. In SQL Server 2005, queries that touch more than one partition use one thread (and thus one processor core) per partition. This sometimes limits the performance of queries that involve partitioned tables, especially when running on parallel shared memory multiprocessor (SMP) computers with many processor cores. Partitioned table parallelism improves the performance of parallel query plans against partitioned tables by better utilizing the processing power of the existing hardware, regardless of how many partitions a query touches. The feature works by default without the need for manual tuning or configuration.

Partition-Aligned Indexed Views

Partition-aligned indexed views enable you to create and manage summary aggregates in your relational data warehouse more efficiently, and use them in scenarios where you couldn’t effectively use them before, improving query performance. In a typical scenario, you have a fact table that is partitioned by date. Indexed views (summary aggregates) are defined on this table to help speed up queries. When you switch in a new table partition, the matching partitions of the partition-aligned indexed views defined on the partitioned table switch too, and do so automatically.

This is a significant improvement over SQL Server 2005

GROUPING SETS

GROUPING SETS allow you to write one query that produces multiple groupings and returns a single result set.

MERGE

The MERGE statement allows you to perform multiple Database Manipulation Language (DML) operations (INSERT, UPDATE, and DELETE) on a table or view in a single Transact-SQL statement.

T-SQL Tuesday #009: Beach Time

I am actually able to get away. Based on the fact that I am digging out from the post-vacation pile you be the judge if I am successful at getting away without creating problems for myself later… but given the fact that I am taking a time-out from all that to write this post, I think I have managed.

There are several things that come to mind…. Things need to work while you are out, the company needs to function (you still need your job to pay for vacation when you get back) thus my need to bend the rule about no technology… if something really bad happens, somehow, some way, I still need to be accessible, and be able to react (maybe not as quick as if I am at work… but no contact is not generally great). For this I have my cell-phone, as well as a cellular wireless card for my netbook. In general, I have alerts, and monitors that call my phone (I want to know there are issues before someone calls me… vacation or not). If something bad happens to a server or database, I will know it…. if it comes back up clean, I will know it…. As for phone-calls, I will accept calls from my boss. He knows I am on vacation, and respects the need to get away, Other than that, I will not answer. As for the netbook, I try to check on my email once a day (maybe two days), I find it much more relaxing to come back to the office, and have all the system messages, etc already cleared from my inbox, and have some idea of the issues that are hot that need my immediate attention. Showing back up to the office with a hundreds of messages, and then having to sort to find the issues is far worse (in my view) than staying on top of it in the first place. Out of office is set…. I do not respond to meeting requests, questions, etc I am on vacation after all. If I have a message from my boss (the one that respects my need to get away), I will respond… the business needs to run, and if a simple email response can keep things on-track in my absence then OK.

So I guess that gets me to my personal rules.

1. Set out of office notices on the phone, and email.
2. Don’t answer calls
   1. Do check messages on the cell phone
   2. Do return a call if it really is necessary
3. Don’t answer emails
   1. Meeting requests… they can wait (people don’t need to know your checking email)
   2. Questions…. They can wait (unless really necessary… in which case I provide enough info for my boss to answer the question)
4. Turn OFF tweetdeck … I love you guys…. But when get away, I need to disconnect, and get away….
5. Don’t tell people you are checking email, and can be reached… the person that needs to know… knows… everyone else, you are on vacation.

Now for the before I go…

I am a Data Warehouse Administrator…. So there are some DBA tasks, and some Load tasks etc.

1. Space seems to be one of those battles I am always fighting…. Make sure there is enough!! The LAST thing you want is full data files while you are out, and load that don’t work for a perfectly preventable reason. Yes I have monitors for these sorts of things, by why mess with that while you are out.
2. Ensure the index maintenance jobs are caught-up, and things are working the way they should from a performance perspective…. If something unfortunate should happen, and you miss index maintenance for a week… you will most likely be OK (assuming your index’s are in good shape before you go).
3. Make sure backups are complete… and up-to-date. Nobody wants to miss backups, but at least be sure you are in good shape before you go.
4. If there are manual daily tasks that must be performed, make sure the person who is going to be doing them for you know exactly what to do… and what to do with the results of those tasks.
5. If at all possible… 4 days before vacation…. NOTHING new goes into production. You want to have some assurance that loads are stable, performance is working… the only thing worse than a change causing problems, is that last minute “make things better” change that shoots you in your own foot.

Hopefully you too can get time away from the office…. get the much needed re-charge you need, and still have a stable working environment upon your return!

This topic seems so simple, and so obvious to me, and do not mean to insult anyone’s intelligence.  However, it is shocking to me the number of people that I have talked to (running large databases) that are not taking advantage  of one of the best performance, and manageability enhancements out there!  Think about the classic phone-book example everyone uses for index’s…. but broaden the example a little.  Consider you have a phone book from NY City, LA, Miami, and Chicago all combined into a single book…. even well indexed it is going to a little bit of effort to find all of the records for a given name in a given city.  Now if you partition that giant book on city, you are not even going to go look in the books that you’re not interested in. …OR… lets say you wanted to find a given name in all four cities, if you have one giant book… you will start looking… if you have the book partitioned into separate books, you could hand one book to three of your friends (threads) and each of you start looking in parallel… clearly divide and concur will get you the answer much more quickly.  

 In a warehouse environment I suggest finding some sort of logical annual partition.  It may be a calendar year, it may be a fiscal year, it may be some annual business cycle.  Make sure what you pick will make sense to your users (to make use if this power they need to include it as filter/join criteria in their queries).  If you pick an annual partition based on calendar year, however the business data is always looked at based on a fiscal year ending in July, even if the queries are well written to filter on the year, you will always use 2 partitions to answer a question because the fiscal year splits 2 calendar years.  

Example

 You can use the wizard in SSMS, or scripts either way you will need to create a partition function  

CREATE PARTITION FUNCTION [year_PF1](INT)
   AS range LEFT
   FOR VALUES (2005, 2006, 2007, 2008,
               2009, 2010, 2011, 2012,
               2013, 2014)  

Once you have the partition function, you will need the partition scheme to be associated to the files groups using a partition scheme  

CREATE PARTITION SCHEME [year_PS1]
   AS PARTITION [year_PF1] TO
   ([data_2005], [data_2006],
    [data_2007], [data_2008],
    [data_2009], [data_2010],
    [data_2011], [data_2012],
    [data_2013], [data_2014],
    [data_2014])  

Now you can create a table on the partition scheme allowing all data to be grouped together by year. The key portion of this statement is the “on” at the end of the table create statement.  

CREATE TABLE partition_example
   (      
      example_year  INT,
       fname         VARCHAR(100),
       lname         VARCHAR(100),
       rec_create_dt DATETIME,
       rec_mod_dt    DATETIME,
       audit_key     INT
   )  ON year_ps1(example_year)  

Conclusion

This is an overly simple example, however, please do yourself the favor and understand how partitioning works.  See my prior discussions on disk management to see how you can simply move data files from one drive to another, as well as the great optimization, and parallelism performance improvements you will see.

Speaker Rate Me!

This is a FULL hour, normally a 75 minute presentation, so be prepared for a lot of information at a quick pace!  I hope you can join me for some layout, recommendations, and ways to look at your disk IO with something other than “it depends”.

I have written in the past about setting up the resource governor, which I am counting as my favorite new(ish) feature, so I am taking this TSQL2sday post opportunity do my promised followup for monitoring of the resource governor.

Introduction

To start with, my advice about the resource governor is to start small.  Have some idea of what you want when you are done… do you need three pools, five pools… how may groups do you envision etc.  Once you “think” you know where you want to go, start small.  Start with a basic A/B split (user interactive Vs Batch process for example) and work your way up from there, monitoring the whole way.  Well balanced, the resource governor works very well… all your resources being consumed in a single pool gets you nowhere (and depending on your max settings can set you back quite a bit)…. so to achieve the “well balanced” lets look at monitoring.

The primary method for monitoring the internals of the resource governor is through the Performance Monitor statistics made available.

Through these metrics you will be able to see both the pool and group level information in much the same way that you would normally see sql server performance statistics, however they are now broken down at a lower level (so you can see how your allocations are performing).  Keep in mind an important point about resource governor, if you allocate 10% of the CPU as a max to a pool, that pool can still get 100% of the CPU if nothing else is running on the machine.  The resource governor ONLY “kicks in”, holds back resources, splits resources among pools (however you wish to think of it) if there is competition between pools.  This makes it important to determine how many, and which resources you are going to assign to the pools.  Pool Stats:

I have setup

• Default – PLEASE don’t get rid of the default pool… I had discussion of this in my prior resource governor post.
• Internal – Internal DB processes run here
• adhoc_pool_high – This is where I have my “normal” interactive user queries, reports.
• adhoc_pool_mde – This is where I have the “High Utilization” tools such as SPSS.  There are tools that will select large amounts of data for mining or other analysis where runtime is not measured in seconds, and can be separated to prevent those very large queries from dominating the machine while still running in reasonable times.
• load_pool_high – This is where I have batch processes.  There is a scheduled job that changes priorities for this pool to a higher-level overnight for my batch-load processes.
• maintenance_pool_low – This is where I run long-running maintenance processes that I just need to get done, but don’t care how long it takes… compressing historical data etc.  Things I want to keep out of the way of the rest of the system.

As you can see in the screenshot there is little else running right at this moment, other than the adhoc_pool_high, so in reality that pool has access to 100% of the cpu until there is something that shows up in one of the other pools that needs resources.

Workgroup setups

Assigned to each of the pools contains one or more workgroups, in this case you can see that the 74% of the adhoc_pool_high is all being used by the “special_user” workgroup.  This is where starting simple will help you greatly, there is no association (other than your knowledge… or looking in the database) between workgroups and pools that you can see in performance monitor.

My associations are

Default pool
   Default workgroup
Internal pool
   internal workgroup
Adhoc_pool_high
   Special_user (specific ID’s that I wish to GUARANTEE performance to)
   dashboard_user (Interactive BI dashboard that NEEDS to perform well)
   User Interactive (interactive reports, management studio, etc_
Adhoc_pool_med
   report_user (scheduled reports)
   user_tools (General bucket for tools other than interactive queries)
   user_spss (SPSS gets it own resource allocation to prevent it from overrunning the machine)
Load_Pool_High
   load_batch
Maintenance_Pool_low
   maintenance

It is at this level where you will be monitoring the resource allocations to see that they are meeting your expectations.  Resource governor is really where art and science meet, you will need to determine which workgroups/pools you wish to give priority to, and which ones will take the back-seat.  Once things are running that cross between workgroups that are assigned to different pools you will see the cpu control % numbers take effect.

In addition to the performance monitor stat, which I have personally found most useful for immediate feedback on the performance of the pools there are a few DMV’s that collect information along the way.  It is through these DMV’s that you can understand the associations between the workload group and the associated pool, as well as what your configurations are.

dm_resource_governor_workload_groups  Where you can see what your allocations/configurations are.
dm_resource_governor_resource_pools Where you can see what your allocations/configurations are.
dm_resource_governor_resource_configuration Is it enabled?

I have been to plenty of conferences over the years, but never having never attended codecamp, I was not exactly sure what I was walking into.  In addition, never having been at a codecamp, having to present to a group of distinguished peers, I was not exactly sure what I was walking into.  What I walked into, was a group of welcoming, helpful, friends that were… to a person… ready to accept a new friend into their group.  Not only was my presentation well received, I feel that I have made some connections that I can count among my friends. … not the linked-in “friend” (that you don’t even know)… but a great group of people with a common goal of providing outstanding training to their peers, just because they can.

The event started with the Friday evening speakers’ dinner, which as you can imagine, was a high-power, room full of geeks (I am including myself in that geek designation… so nobody take that badly!) … a good time was had by all.

Saturday (in what I can call a truly impressive facility at J. Sergent Reynolds Community College) we kicked off the event, with a few hundred folks ready to go by 7:30 AM on a Saturday morning.. and a few hundred more by the time the first sessions started.  A few sessions into the morning, the only thing I can say…. is I was getting a bit nervous…. I had given the presentation before, but the presenters I saw were nothing less than professional speakers that knew their area of expertise so well it was starting to scare me a bit. Not to mention the fact that there were a few hundred really smart people willing to give up a Saturday to learn something. OK… I have the after-lunch session, so maybe people will be full and sleepy… is that good?!? not sure… Sleeping audience will not ask questions… but a sleeping audience is not exactly motivational…

Turns out the presentation was well timed (a TON of information to pack into just over an hour), reasonably well attended, and very well received.  Since it was the after lunch session, I was able to setup early and was answering questions from folks right up to the start-time of the presentation… so my nerves never had a chance to even enter into the picture.  I honestly can’t say I remember exactly how the hour+ went by in about 45 seconds, but in the end, I made it through the slides and demo with about 3 minutes to spare for questions.

The time-warp seemed to end after my presentation…and I felt good that (even those that attended this presentation for a second time now) stopped me and expressed how helpful it was, and how good my delivery of the material is… no I am not trying to pat myself on my own back on my own blogg, but seriously there were some GREAT speakers at codecamp… if people are comparing me to those folks and liking what I had to say… they (probably ate too much lunch)… so I guess the after lunch time slot worked to my advantage!

I tip my hat to the hard work of all the organizers… my new (and old) friends for the support! Unfortunately for those that will sit through my presentations in the future, I will be back for more!!

If you came out to see this presentation, please provide me some feedback through speaker-rate.  The feedback helps me build better presentations… and hopefully provide better, more informative sessions to you in the future.

I enjoyed the whole codecamp experience, and I hope that you got as much out of the day as I did!!  A truly impressive group of presenters with a great breadth of topics, I was glad to be a small part of the great day.

As I mentioned, everything I referenced in the presentation can be found my series on disk performance:

Help — My disk performance      (RAID how/why what to put where)
Help — My disk performance II  (Physical connections to the SAN)
Help — My disk performance III  (Partitioning and compression)
Help — My disk performance IV (filegroups and files)  
Help — My disk performance V   (Monitoring)  
Help — My disk performance VI (Additional monitoring and conclusions)

Also, as promised, the presentation is available for reference on my Presentations page