Ryan Adams Blog

SQL, Active Directory, Scripting

SQLSaturday Dallas

The North Texas SQL Server User Group Board of Directors has decided to do something a little bit different for selecting Precons for our next SQLSaturday.  The SQLSaturday event will be held on October 3rd with the precons on October 2nd.  We have decided that since it is the community and our user group members that attend these precons that they should choose who gets selected.  Our leadership is constantly looking for ways to stay rooted in the community and give back so we’re excited since we think this does exactly that.  If you are paying money to attend these valuable full day training sessions shouldn’t you have some say in what the topics are?  Well now you do!

We are going to open the call for Precon speakers from August 1st to August 14th.  If you want to submit, you just need to send your Bio and Abstract to precons@ntssug.com.  We’ll post all of the submissions HERE by August 15th and the voting will begin from there until August 21st.  In order to vote, Click Here to send an email with the precon title as your subject, between August 15th and 21st.  You are only allowed to vote for 1 precon since you can only be in one place at a time, so vote for the one you would want to attend.  We’ll tally the votes for the top DBA session, BI session, and a wildcard.  The winners will be announced by August 24th.

The precons will be held on October 2nd at the Microsoft offices in Irving.  7000 North Highway 161 Irving, Texas 75039 Building LC-1.  We’ll have an early bird price of $99 and the full price will be $125.

SQLSAT422_web

I’ll be doing a precon for SQLSaturday #422 in Oklahoma City on August 28th, 2015 titled “Creating a High Availability and Disaster Recovery Plan” .  If you are in the area or thinking of attending this event, I would love to have you in the class.  We are going to cover backups, windows clustering, AlwaysOn Failover Cluster Instances, AlwaysOn Availability Groups, and more.  This class will take each of those technologies in a progressive order to build on each other.  At the end of the day we will have a single solution built out in virtual machines on my laptop that use all of those technologies to build a comprehensive and real world high availability and disaster recovery plan.

Still not convinced?  Here are some testimonials from previous classes.

Denver Attendee

This past weekend was SQL Saturday #331 Denver. I always enjoy SQL Saturdays, but I was especially looking forward to this weekend because I was going to attend a SQL Saturday Pre-Con for the first time! The pre-con was fantastic! The Denver SQL Server User Group did an outstanding job, and it was a wonderful day of learning and networking.

I attended Ryan Adams’ session A Day of High Availability and Disaster Recovery. Ryan did an excellent job with this all-day session. You could tell that his agenda was carefully planned to build upon each module. Ryan’s first module was an in-depth tour of backups and restores, pointing out that these concepts were foundational to all High Availability (HA) and Disaster Recovery (DR) solutions. In his next module, Ryan expanded on backup/restores as he introduced us to mirroring. Ryan then walked us through the concept of Windows Server Failover Clustering (WSFC). Culminating in his module on AlwaysOn Failover Cluster Instances and Availability Groups, Ryan tied all the concepts together providing a holistic view of HA/DR. I would definitely recommend this all-day session to anyone looking to expand his or her knowledge of HA/DR. To find out more about Ryan Adams, you should check out his blog at ryanjadams.com.

Madison Attendee

One thing I liked the most about this session is the speed of the presenter was just right.  He would get our feedback frequently to see how fast or slow we want to move.  He had a lot of real world examples from his own experience and from other participants.

Madison Attendee

Great lecture!  Ryan is a very logical presenter; made the material extremely easy to understand.  I greatly enjoyed the class.

Want the full description?  Here it is:

Abstract: You have been tasked to configure a high availability and/or disaster recovery plan to ensure your company’s systems are always up and running while able to withstand a disaster. SQL Server offers several options, but how do you choose the right one to meet your business requirements? I’ve been faced with this same problem. I’ll help you figure out how to choose the right technology. We will start off with a solid foundation in backups and demonstrate how to configure them for initializing the various HA/DR technologies while also cutting your backup and restore time in half. You will see how you can keep an offsite copy of your database in sync with your production system with mirroring.

You’ve read that several of the HA/DR solutions require Windows Failover Clustering, but you don’t know anything about managing a complicated Windows cluster. I’ll show you how to setup and configure a Windows cluster and we’ll build on that with SQL Server AlwaysOn Failover Clustering Instances that provide the 99.999% of up-time your management demands. Lastly we’ll dive into how to setup AlwaysOn Availability Groups to solve both HA and DR in one solution. We will see how all these technologies work together by discussing a case study and developing a comprehensive solution just like you have to do when you return to the office. Here’s what you will learn:

Planning a backup strategy to avoid a data disaster
Cutting your backup and restore time in half when initializing DBs for Mirroring or Availability Groups
Configuring mirroring to keep production in real-time sync with an offsite database copy
Learning configuration tips to increase throughput
Configuring a Windows Failover Cluster to meet that five 9s business requirement
Discovering how to properly configure a Windows Cluster quorum to support the AlwaysOn feature set
Implementing a SQL AlwaysOn Failover Cluster Instance
Creating a SQL AlwaysOn Availability Group to solve both your HA and DR requirements in one solution

You’re just as excited and convinced as these fine folks, right? Then click the link below to get signed up.

Eventbrite_Button

PASS_VC_PerformanceHow could you possibly argue with getting 8 straight hours of SQL Server Performance training for FREE?

You can’t!  Go check out the annual Performance Palooza hosted by the Performance Virtual Chapter on July 23rd, 2015.  This is a virtual online event open to the public and it’s absolutely free.  We’ll have 8 back to back sessions with start times from 10am to 5pm Central Time.

How are we able to provide all of this for free?  We want to thank Solarwinds for sponsoring our chapter and this event so that we can bring it to you completely free!

In fact Solarwinds was so generous that we will be raffling a $50 Amazon gift card to a random attendee out of every single session.

Thanks Solarwinds!  SolarWind

So what can you learn?  Here is the list of sessions and the links to register for each of them.

Central TimeSpeakerTopic
10amJohn SterrettMake Your SQL Server Queries Go Faster
11amWarner ChavesBuilding High Performance SQL Servers Virtual Machines on AWS and Azure
12pmAndy GalbraithPerforming a SQL Server Health Check
1pmMatan YungmanColumstore Indexes – Questions and Answers
2pmNeil HamblyPerformance Troubleshooting Using DMVs
3pmTim MitchellMaximizing SSIS Package Performance
4pmTrayce JordanTroubleshooting Seconday Replica Latency
5pmMike FalSQL Server Benchmarking: The Powershell Speedometer

I’ll be doing a precon for SQLSaturday #421 in Columbus Ohio on July 10th, 2015 titled “Planning a High Availability and Disaster Recovery Plan” .  If you are in the area or thinking of attending this event, I would love to have you in the class.  We are going to cover backups, windows clustering, AlwaysOn Failover Cluster Instances, AlwaysOn Availability Groups, and more.  This class will take each of those technologies in a progressive order to build on each other.  At the end of the day we will have a single solution built out in virtual machines on my laptop that use all of those technologies to build a comprehensive and real world high availability and disaster recovery plan.

Still not convinced?  Here are some testimonials from previous classes.

Denver Attendee

This past weekend was SQL Saturday #331 Denver. I always enjoy SQL Saturdays, but I was especially looking forward to this weekend because I was going to attend a SQL Saturday Pre-Con for the first time! The pre-con was fantastic! The Denver SQL Server User Group did an outstanding job, and it was a wonderful day of learning and networking.

I attended Ryan Adams’ session A Day of High Availability and Disaster Recovery. Ryan did an excellent job with this all-day session. You could tell that his agenda was carefully planned to build upon each module. Ryan’s first module was an in-depth tour of backups and restores, pointing out that these concepts were foundational to all High Availability (HA) and Disaster Recovery (DR) solutions. In his next module, Ryan expanded on backup/restores as he introduced us to mirroring. Ryan then walked us through the concept of Windows Server Failover Clustering (WSFC). Culminating in his module on AlwaysOn Failover Cluster Instances and Availability Groups, Ryan tied all the concepts together providing a holistic view of HA/DR. I would definitely recommend this all-day session to anyone looking to expand his or her knowledge of HA/DR. To find out more about Ryan Adams, you should check out his blog at ryanjadams.com.

Madison Attendee

One thing I liked the most about this session is the speed of the presenter was just right.  He would get our feedback frequently to see how fast or slow we want to move.  He had a lot of real world examples from his own experience and from other participants.

Madison Attendee

Great lecture!  Ryan is a very logical presenter; made the material extremely easy to understand.  I greatly enjoyed the class.

Want the full description?  Here it is:

Abstract: You have been tasked to configure a high availability and/or disaster recovery plan to ensure your company’s systems are always up and running while able to withstand a disaster. SQL Server offers several options, but how do you choose the right one to meet your business requirements? I’ve been faced with this same problem. I’ll help you figure out how to choose the right technology. We will start off with a solid foundation in backups and demonstrate how to configure them for initializing the various HA/DR technologies while also cutting your backup and restore time in half. You will see how you can keep an offsite copy of your database in sync with your production system with mirroring.

You’ve read that several of the HA/DR solutions require Windows Failover Clustering, but you don’t know anything about managing a complicated Windows cluster. I’ll show you how to setup and configure a Windows cluster and we’ll build on that with SQL Server AlwaysOn Failover Clustering Instances that provide the 99.999% of up-time your management demands. Lastly we’ll dive into how to setup AlwaysOn Availability Groups to solve both HA and DR in one solution. We will see how all these technologies work together by discussing a case study and developing a comprehensive solution just like you have to do when you return to the office. Here’s what you will learn:

Planning a backup strategy to avoid a data disaster
Cutting your backup and restore time in half when initializing DBs for Mirroring or Availability Groups
Configuring mirroring to keep production in real-time sync with an offsite database copy
Learning configuration tips to increase throughput
Configuring a Windows Failover Cluster to meet that five 9s business requirement
Discovering how to properly configure a Windows Cluster quorum to support the AlwaysOn feature set
Implementing a SQL AlwaysOn Failover Cluster Instance
Creating a SQL AlwaysOn Availability Group to solve both your HA and DR requirements in one solution

You’re just as excited and convinced as these fine folks, right? Then click the link below to get signed up.

Eventbrite_Button

SQLSAT_Puerto Rico

I’ll be doing a precon for SQLSaturday #373 in Puerto Rico on May 8th, 2015 titled “Planning a High Availability and Disaster Recovery Plan” .  If you are in the area or thinking of attending this event, I would love to have you in the class.  We are going to cover backups, windows clustering, AlwaysOn Failover Cluster Instances, AlwaysOn Availability Groups, and more.  This class will take each of those technologies in a progressive order to build on each other.  At the end of the day we will have a single solution built out in virtual machines on my laptop that use all of those technologies to build a comprehensive and real world high availability and disaster recovery plan.

Still not convinced?  Here are some testimonials from previous classes.

Denver Attendee

This past weekend was SQL Saturday #331 Denver. I always enjoy SQL Saturdays, but I was especially looking forward to this weekend because I was going to attend a SQL Saturday Pre-Con for the first time! The pre-con was fantastic! The Denver SQL Server User Group did an outstanding job, and it was a wonderful day of learning and networking.

I attended Ryan Adams’ session A Day of High Availability and Disaster Recovery. Ryan did an excellent job with this all-day session. You could tell that his agenda was carefully planned to build upon each module. Ryan’s first module was an in-depth tour of backups and restores, pointing out that these concepts were foundational to all High Availability (HA) and Disaster Recovery (DR) solutions. In his next module, Ryan expanded on backup/restores as he introduced us to mirroring. Ryan then walked us through the concept of Windows Server Failover Clustering (WSFC). Culminating in his module on AlwaysOn Failover Cluster Instances and Availability Groups, Ryan tied all the concepts together providing a holistic view of HA/DR. I would definitely recommend this all-day session to anyone looking to expand his or her knowledge of HA/DR. To find out more about Ryan Adams, you should check out his blog at ryanjadams.com.

Madison Attendee

One thing I liked the most about this session is the speed of the presenter was just right.  He would get our feedback frequently to see how fast or slow we want to move.  He had a lot of real world examples from his own experience and from other participants.

Madison Attendee

Great lecture!  Ryan is a very logical presenter; made the material extremely easy to understand.  I greatly enjoyed the class.

Want the full description?  Here it is:

Abstract: You have been tasked to configure a high availability and/or disaster recovery plan to ensure your company’s systems are always up and running while able to withstand a disaster. SQL Server offers several options, but how do you choose the right one to meet your business requirements? I’ve been faced with this same problem. I’ll help you figure out how to choose the right technology. We will start off with a solid foundation in backups and demonstrate how to configure them for initializing the various HA/DR technologies while also cutting your backup and restore time in half. You will see how you can keep an offsite copy of your database in sync with your production system with mirroring.

You’ve read that several of the HA/DR solutions require Windows Failover Clustering, but you don’t know anything about managing a complicated Windows cluster. I’ll show you how to setup and configure a Windows cluster and we’ll build on that with SQL Server AlwaysOn Failover Clustering Instances that provide the 99.999% of up-time your management demands. Lastly we’ll dive into how to setup AlwaysOn Availability Groups to solve both HA and DR in one solution. We will see how all these technologies work together by discussing a case study and developing a comprehensive solution just like you have to do when you return to the office. Here’s what you will learn:

Planning a backup strategy to avoid a data disaster
Cutting your backup and restore time in half when initializing DBs for Mirroring or Availability Groups
Configuring mirroring to keep production in real-time sync with an offsite database copy
Learning configuration tips to increase throughput
Configuring a Windows Failover Cluster to meet that five 9s business requirement
Discovering how to properly configure a Windows Cluster quorum to support the AlwaysOn feature set
Implementing a SQL AlwaysOn Failover Cluster Instance
Creating a SQL AlwaysOn Availability Group to solve both your HA and DR requirements in one solution

You’re just as excited and convinced as these fine folks, right? Then click the link below to get signed up.

Eventbrite_Button

SQLSaturday #387 Madison

I’ll be doing a precon for SQLSaturday #387 in Madison Wisconsin on April 10th, 2015 titled “Planning a High Availability and Disaster Recovery Plan” .  If you are in the area or thinking of attending this event, I would love to have you in the class.  We are going to cover backups, windows clustering, AlwaysOn Failover Cluster Instances, AlwaysOn Availability Groups, and more.  This class will take each of those technologies in a progressive order to build on each other.  At the end of the day we will have a single solution built out in virtual machines on my laptop that use all of those technologies to build a comprehensive and real world high availability and disaster recovery plan.

Still not convinced?  Here is a testimonial from a previous class.

This past weekend was SQL Saturday #331 Denver. I always enjoy SQL Saturdays, but I was especially looking forward to this weekend because I was going to attend a SQL Saturday Pre-Con for the first time! The pre-con was fantastic! The Denver SQL Server User Group did an outstanding job, and it was a wonderful day of learning and networking.

I attended Ryan Adams’ session A Day of High Availability and Disaster Recovery. Ryan did an excellent job with this all-day session. You could tell that his agenda was carefully planned to build upon each module. Ryan’s first module was an in-depth tour of backups and restores, pointing out that these concepts were foundational to all High Availability (HA) and Disaster Recovery (DR) solutions. In his next module, Ryan expanded on backup/restores as he introduced us to mirroring. Ryan then walked us through the concept of Windows Server Failover Clustering (WSFC). Culminating in his module on AlwaysOn Failover Cluster Instances and Availability Groups, Ryan tied all the concepts together providing a holistic view of HA/DR. I would definitely recommend this all-day session to anyone looking to expand his or her knowledge of HA/DR. To find out more about Ryan Adams, you should check out his blog at ryanjadams.com.

Want the full description?  Here it is:

Abstract: You have been tasked to configure a high availability and/or disaster recovery plan to ensure your company’s systems are always up and running while able to withstand a disaster. SQL Server offers several options, but how do you choose the right one to meet your business requirements? I’ve been faced with this same problem. I’ll help you figure out how to choose the right technology. We will start off with a solid foundation in backups and demonstrate how to configure them for initializing the various HA/DR technologies while also cutting your backup and restore time in half. You will see how you can keep an offsite copy of your database in sync with your production system with mirroring.

You’ve read that several of the HA/DR solutions require Windows Failover Clustering, but you don’t know anything about managing a complicated Windows cluster. I’ll show you how to setup and configure a Windows cluster and we’ll build on that with SQL Server AlwaysOn Failover Clustering Instances that provide the 99.999% of up-time your management demands. Lastly we’ll dive into how to setup AlwaysOn Availability Groups to solve both HA and DR in one solution. We will see how all these technologies work together by discussing a case study and developing a comprehensive solution just like you have to do when you return to the office. Here’s what you will learn:
Planning a backup strategy to avoid a data disaster
Cutting your backup and restore time in half when initializing DBs for Mirroring or Availability Groups
Configuring mirroring to keep production in real-time sync with an offsite database copy
Learning configuration tips to increase throughput
Configuring a Windows Failover Cluster to meet that five 9s business requirement
Discovering how to properly configure a Windows Cluster quorum to support the AlwaysOn feature set
Implementing a SQL AlwaysOn Failover Cluster Instance
Creating a SQL AlwaysOn Availability Group to solve both your HA and DR requirements in one solution

You’re just as excited and convinced as these fine folks, right? Then click the link below to get signed up.

Eventbrite_Button
SQLSatMadisonTweets

 

 

 

 

We are back in the keynote room with Adam Jorgensen talking to us about the financial state and health of the PASS organization.  It’s currently very healthy with over $1 Million in reserve.

Thomas LaRock is back and thanking those folks who are rolling off the PASS board and have dedicated so much time and PASSion.  Sri Sridharan is thanked for his amazing work in helping volunteers take on bigger challenges and being successful.

Denise McInerney talks about how she got involved with the PASS community and her journey of how it helped her become so successful in her career.  This leads up to the awarding of the 2014 PASSion award.  There is only 1 of these given out every year to a volunteer who has gone even more than above and beyond for the organization.  Andrey Korshikov from Russia is announced as the winner.  Congrats Andrey!

Denise announces the PASS Business Analytics Conference and also the dates for the next PASS Summit back here in Seattle.

Dr. Rimma Nehme is coming on stage to talk about Cloud Databases.  She talks about where she is from and how she got where she is as MS Grey Labs today.  She is explaining what cloud computing is, as a computing service that is available from anywhere, at anytime, and is always on. Cloud computing is an on demand service, location transparent, and rapid delivery.  First she is explaining cloud elasticity that features quick and easy deployment.  She is showing some very cool pictures on Microsofts physical cloud infrastructure (datacenter).  Everything is stored in shipping containers (18 wheeler trailers).  She shows all the servers and how they look on the inside of the container.  A very cool look behind the scenes.

Now she is explaining how they get all the power to these DCs that they need and how they cool them with “swamp cooling”. She is now explaining software as service using the analogy of pizza as a service.  A great and easy to remember explanation.

Dr. Nehme moves on to explaining virtualization and the cloud.  The analogy here is roommates having to share a bathroom with a single sink and how now get a large sink to service everyone.  We now move on to multi-tenancy and the tiered approach and options. She now explains SLAs. 3 nines gives you 8.76 hours of down time a year and 4 nines gives you 52.56 minutes of down time.

SQL DB in Azure is explained as we look at it from the software perspective.  We have an the infrastructure layer (hardware), platform layer, services layer, and client layer.

Dr. Nehme is now talking about how we still need DBAs even with all the automation and simplicity of the cloud.  She encourages us take our skills, add cloud, and give ourselves a new title as “Cloud DBA”.

Okay folks that’s it for today!

 

This morning I will be live blogging the PASS Summit 2014 Day 1 keynote so those at home can follow one.

Thomas LaRock hits the stage and welcomes everyone to the Summit.  He thanks Microsoft as the makers of the best data platform on the planet along with all the volunteers and leaders in this wonderful community.  He is giving us a “state of union” of PASS with amount in chapters, virtual chapters, and training hours provided.

2oo hands-on workshops are being taught and provided by Microsoft and the ability to take certification exams are available this week.  The SQL Server Clinic is also available which is an area staffed by Microsoft Customer Service and Support and where you can get your technical issues addressed directly from Microsoft.

This morning’s keynote is being presented by Ranga Rengarajan, James Phillips, and Joseph Siroch from the Microsoft Data Platform.  We are now watching an introductory video where these 3 executives are explaining their roles in the MS Data Platform group and what areas they support.

Ranga says he has 2 daughters and tells them he is getting the industry ready for them as he encourages them toward the tech industry.  Ranga is talking about how he likes maps and how the map availability has changed over time due to data.  He is talking about how data drives so many things in our lives and how things have changed because of it.  Ranga shows all the things that the Microsoft Data Platform can handle all in 1 engine.  This includes things like scale out, scale up, in memory, on disk, OLTP, and data warehouse.  Some of the other aspects he points out are capturing diverse data with Azure DocumentDB and Hadoop.  We can achieve elastic scale with SQL 2014 scaling to 640 logical cores and up to 1TB of RAM per VM, SQL Server on virtual machines in Azure, and Azure SQL DB.

Ranga welcomes Tracy Daugherty from Pier 1 imports to show how they use some of the new Azure services for their business.  His demo is showing how they are trying to sell decorative pumpkins during the holiday season and how they can dynamically display seasonal items based on inventory.  This lets them know what they need to push from a sales perspective and how Azure allows them to do all this dynamically and without code changes.  He is now showing us how they can use elastic scale to handle the additional load during the seasonal promotions.  The last thing he is showing is Azure Geo Replication and how easy it is to add a replica across the country.

The reigns are handed back over to Ranga as he shows some customers that are using some scale out features.  He features StackOverflow and some of the interesting things that they have done and we are now watching a video about how Dell uses various features to scale out and and scale up their infrastructure.  In memory has had a big impact for them and they are happy with the ease of deployment in SQL 2014 and how their workloads have performed better on the platform without any changes.

Now that we have seen everything MS has done Ranga is going to talk about what they are doing in the future.  They have a major update coming to Azure SQL DB including many larger indexes, parallel queries, extended events, and in memory.  Unfortunately with the announcement the crowd is silent.

Ranga welcomes Mike Zwilling to the stage also from Pier 1 Imports.  He shows how you can now use in memory and columnstore at the same time on the same table.  Mike is now showing how you can take an on-premise table and stretch it into Azure.  We see how you can still query the data while it is being migrated.  He is simulating a crash of the local server and how long it takes to restore a 150GB database which only takes seconds and then synchronizes back with the data in Azure and maintaining transactional consistency.

Joseph Sirosh is welcomed to the stage, VP of machine learning and information management.  He is going to have the audience do the wave while shouting PASS…Community…Rocks.  He is talking about the Azure machine learning platform and how we can use the data from machines to analyze and make predictions.  We are seeing how Pier 1 Imports can use MS Kinect as a sensor to  see where customers are spending time in their stores.  They are using Azure to show how they are pulling the data in.  They now basically have SSIS in cloud and through the browser where you can pull from multiple data sources.  We are now seeing the real time data coming from the Kinect sensors through Azure streaming and it appears that candles are the hot item of the day.  Sanjay has his phone plugged in so we can hear the real time predictions for his Pier 1 Shopping experience.  He now asks his phone to find cocktail glasses and the application finds the product and tells us what department it is in.  It also brings up a map to show us where to find it in the store.

James Phillips the GM of Data Experiences comes out to tell us about some of the Power BI offerings.

I’m going to close it out as I head out to talk about AlwaysOn in room 6E, which will be live streamed on PASStv.  If you are here then I hope you’ll join me and if not I hope you’ll join me online at 10:15 PT.

I’ll be back to live blog tomorrow’s keynote as well.

80s Tees Tommy Boy T-Shirt

I was just awarded as a Microsoft SQL Server MVP!

I can’t believe Microsoft has chosen me for this award and to be included in such a prestigious program.  There are only a little over 70 SQL Server MVPs in the United States and a little over 350 world wide.

For my readers who are unfamiliar with the program here is how Microsoft describes it.

Microsoft Most Valuable Professionals, or MVPs are exceptional community leaders who actively share their high-quality, real-world deep technical expertise with the community and with Microsoft. They are committed to helping others get the most out of their experience with Microsoft products and technologies.

You can learn more about the program HERE

I wouldn’t be in this position if it were not for my understanding and supportive wife.  My deepest gratitude and love to her.

I also have to thank the SQL community for all the support and opportunities.  I love this community, and I look forward to continuing to serve and spread some knowledge.

I presented a session titled “SQL Server AlwaysOn Quickstart” on September 10th for the 24 Hours of PASS event.  You can view the session recording HERE if you missed it.  The session was a preview of the full session I will be presenting at the PASS Summit 2014 this November.  I will be presenting the full session on the first day of the Summit, November 5th, 2014 at 10:15am PST.

The session is right before lunch which means I don’t have to hurry out to make room for the next presenter since there is a 2 hour lunch break.  You came to the Summit to learn about SQL Server and get your questions answered so I’ll stay after the session for as long as it takes to answer every question you have.  If I don’t have the answer then I’ll find it for you and I really mean that.  Ask anyone who I promised to followup with after one of my sessions.

Make sure to add my session when you’re using the Schedule Builder to plan your Summit sessions!  Now on to the questions I received during the 24HOP session.

  • The FCI needs shared storage and the AG needs local storage. I’m confused on how they can be implemented together.
    • The nodes in your cluster can have both shared storage and local storage.  The AG will use the local storage (could still be SAN attached just not shared) and essentially mirror the DBs on all nodes.  The FCI will use the shared storage and might not even be installed across all nodes in the cluster.  They are absolutely supported together.  Come see my session and I’ll show you how.  Click here to download one of my presentations showing you a sample architecture.
  • I noticed you did not add the system databases to your availability group, is there a reason for that?
    • Yes.  Just like mirroring system databases are not eligible and cannot be added to an AG.
  • If SQL 2012 supports 1 primary and 4 secondary and SQL 2014 supports 1 primary and 8 secondary, how many are read/write?
    • Only the primary is read/write.  All secondary replicas are read only and you can disable that functionality or limit it if you want.
  • Are the replicas synchronized?
    • For a replica to be considered “synchronized” it must be in Synchronous Mode.  You can only have a maximum of two replicas in Synchronous Mode.
  • Is it common to use a SQL FCI as a primary replica and another SQL FCI as a secondary replica for protection and still take advantage of AGs for reporting purposes?
    • With the exception of the reporting part this used to be common prior to AGs.  You would have two 2 node FCIs in different data centers and mirror between them.  This gave you HA in both data centers and DR if an entire data center went down.  However, this is not so practical with AGs.  You would be more likely to have 2 replicas in each data center with an AG across them and no FCIs.
  • How would you handle getting stored procedures and SQL Agent jobs onto the secondary replicas?
    • You don’t have to worry about stored procedures since those are stored in the database and will be replicated automatically.  SQL Agent jobs will have to be re-created on the secondary replicas and additional code added to exit gracefully if the replica is currently acting as a secondary.  You’ll want them to run automatically if the replica becomes the primary.  See THIS POST for more information on automating logins and thanks to Robert Davis (Blog|Twitter) for writing the code.
  • How do you handle installing SQL Service packs after AGs have been implemented?
    • First make sure you have backups (That’s always rule #1).  You’ll want to update a secondary replica first, make sure it is in synchronous mode, and then fail over to it.  Now you can upgrade the remaining replicas and fail back to your preferred replica when done.
  • Does SELECT @@SERVERNAME return the listener name or the node name?
    • It returns the node name.  However, older versions of SQL Server (for an FCI) will return the listener name.  The best way to get the node name no matter what version you are on is to SELECT SERVERPROPERTY(‘ComputerNamePhysicalNetBIOS’)
  • Wouldn’t it be possible to have an indexed view on a secondary replica to use with reporting off of that replica?
    • You cannot create database objects on secondary replicas separate from the primary.  You have to create them on the primary and they will sync to the secondary replicas.  Remember that secondary replicas are READ ONLY and can never be written to.
  • With AGs in automatic failover mode how does the client connection timeout have to be configured?
    • Your applications need to be written to retry the connection.  You should also add MultiSubnetFailover=true to the connection string.  You should consider reducing the TTL of the listener name.  I’ll write another blog post explaining this in more detail, but that is the short answer.
  • What is the advantage of having a group of databases in a single AG?
    • If you have several databases that support a single application, if any one of them fails then they all failover together.  If you were mirroring them it was possible that myDB2 could failover to the mirror while myDB1 and myDB3 continued to run on the primary.  That would break the application, but AGs mitigate that risk factor since they stay together within that logical construct.
  • Can you have replicas in different domains?
    • No.  All replicas have to be part of the same Windows Cluster which requires that all nodes be in the same domain.  As a previous AD guy I’ll take this a step further and point out that in most situations things are fine to be in different domains so long as they are trusted somewhere in the forest hierarchy.  That is NOT the case here.
  • All servers in our environment are in a cluster of 8 nodes. Would that be okay ,or do you still need to create another cluster inside that cluster?
    • You cannot create a cluster within a cluster.  In this case you don’t need to do anything since you already have a cluster, you’re one step ahead of the game.