This week, Brent, Angie, Erik, Tara, Jessica, and Richie discuss SSMS issues, security auditing, snapshot replication, SSIS Cache Connection Manager, AlwaysON Availability Groups, deadlocks, and Jessica’s trip to Mexico.
You can register to attend next week’s Office Hours, or subscribe to our podcast to listen on the go.
Office Hours Webcast – 2016-06-01
Jessica Connors: All right, I guess we should be talking about SQL Server.
Erik Darling: Nah.
Brent Ozar: Oh no.
Erik Darling: Boring.
Jessica Connors: That one product.
Erik Darling: Snoozefest.
Brent Ozar: Which is out new today. So ladies and gentlemen, if you’re watching this, SQL Server 2016 is out right now. You can go download it on MSDN or the partner site. There’s places where you can go it. Developer Edition is free so you can go download the latest version right now. As we speak, Management Studio is not out yet but will be coming any moment.
Jessica Connors: That was our first question too: Is 2016 out yet?
Brent Ozar: Dun dun dun.
Jessica Connors: Are you hearing any rumblings on problems with it?[Laughter]
Brent Ozar: We all start laughing. There were a lot of problems with the community previews. For example, SQL Server Management Studio would crash every time I would close it. So I’m really curious. Usually you don’t see stuff quite this buggy as you get close to release. But at the same time, I’m like, well, no one ever goes live with it in production the day it comes out anyway. People are just going to get widespread experience in development, in dev environments, and QA, then they’ll go find bugs hopefully and fix them. Hopefully.
Angie Rudduck: Wait. So I shouldn’t install that in our production servers running everything?
Brent Ozar: Yeah, no. I would take a pass for a week or two. Just let things bake out just a little bit.
Jessica Connors: Let it just wait.
Jessica Connors: Let’s take a question from Dennis, SSMS question. Is there a way to have SSMS format the numbers in the output messages? Not data in the query, like the row count at the bottom?
Tara Kizer: What are you trying to solve here? Because this is a presentation layer issue. Management Studio, it’s just a tool for us to query data, why does the formatting of the output matter? If you have an application you’re developing in .NET, format your data there. The row count at the bottom. No, Management Studio, there isn’t a way to format it. You can change the font and things like that in the tools options but I’m not sure that that’s what you’re asking.
Richie Rump: Is there a way, Brent? Could we format that?
Erik Darling: One thing you can do is if you’re interested in just having commas in is you can cast it as money or convert it to money with a different culture and you can get commas put in. But other than that, I’m not really sure what you’re after so if you’re a little more specific.
Brent Ozar: Well and return it as results. Whatever you’re looking for, return it as results instead of looking at what comes out of SSMS. Then you can format it there as well.
Jessica Connors: Dennis hasn’t replied to us.
Jessica Connors: Let’s go to Ben. He says, “[inaudible 00:02:24 old] to SQL. Hearing rumors about going to the cloud, MS, or Amazon, specifically in terms of security. What are the gotchas and pain points? Security is not our forte.”
Brent Ozar: This is so totally different from on-premises because on-premises you don’t have any security risks at all. No one could possibly access your data. I’m sure it’s locked down tighter than the pope’s poop chute. I mean it is completely secure as all get out. Just me, I’m usually like… Erik says, “Pull my finger.” I would say usually it’s more secure because you don’t go wild and crazy with giving everybody sysadmin. So I just turn it back to people on-premises and go, “So let’s talk about your security. Let’s go take a look at what you got. Everybody is SA. You haven’t changed your password in three years? Yeah, I think you should get out of on-premises. On-premises is probably the worst thing for you.” Nate says, “The pope’s poop chute? Really?” Yes. This is what happens when you work for a small independent company. You can say things like “tighter than the pope’s poop chute.” Probably can’t say that but we’ll find out later.[Laughter]
Angie Rudduck: You’ve already said it at least three times, we’re going to find out. You’re going to get an official letter from the pope.
Brent Ozar: The Vatican, yep.
Angie Rudduck: Yeah.
Brent Ozar: “The pope does not have a poop chute.”[Laughter]
Erik Darling: Going for a world record, most references to the pope’s butt in one webcast.
Angie Rudduck: Stop it.
Brent Ozar: Dad always said that to me, so yeah, there we go. Someone else should probably tackle the next question.
Richie Rump: Yeah, somebody else talk now, please.
Jessica Connors: Brent, I’ll just put him on mute.
Erik Darling: Looser than Brent’s…[Laughter]
Erik Darling: Wallet, wallet, wallet.
Angie Rudduck: Wallet on the company retreat.
Brent Ozar: There we go.
Jessica Connors: I’m glad it’s a short week.
Jessica Connors: Question from J.H. “Would creating a server trigger and emailing our DBA team if someone makes changes to the server role safe? Hard triggers affect performance, but it’s rare in our case that we have server role changes but want to catch it if a network admin puts himself in the sysadmin role without letting us know.”
Tara Kizer: We had security auditing at my last job. I’m not too sure what was used. Well, I think the other DBA who set this all up, he just set up a job and queried for the information. Then the job would run every few minutes I believe and would send the DBA team an alert if anything changed.
Brent Ozar: Yeah, I like that. My first reaction was Extended Events.
Tara Kizer: We had really strict auditing that we had to put in place due to the credit card information. It was encrypted but we had to be very careful with everything.[Erik and Brent speaking at same time]
Brent Ozar: Would you say you had tight security? How tight was security? Go ahead, Erik, I dare you.
Erik Darling: Oh, sorry. I was going to say that you can set up the event data. You got me all flustered now. You can set up event data XML. It’s pretty good for modification triggers like that. It’s not like, you know, if you put triggers on tables and you’re doing massive shifts of data or you know before and after stuff. It’s a pretty lightweight way to just log changes as they happen.
Jessica Connors: Let’s see here. Question from Terry. “Is there a way to set up databases in an AG without doing a backup and restore?”
Erik Darling: Not a good one.
Tara Kizer: No.
Brent Ozar: 2016 there is. 2016 we get direct seeding where we can seed directly from the primary, so starting today you can. But unfortunately, not before today.
Jessica Connors: All right, a security question. This is from Nate regarding security auditing. “Any suggestions on getting some basic setup that tracks and alerts for security changes and schema changes?”
Tara Kizer: I don’t know.
Brent Ozar: I don’t know either. Is there like an Extended Event or something you could hook into?
Tara Kizer: Probably. What we had set up for security would have just been queries, just to query for the information. Look at the system tables and views. For schema changes, I don’t know.
Angie Rudduck: I think somebody set up a simple, “Hey, there’s somebody new in this group” for a security group. I think it was PowerShell at my last place just to like all of a sudden somebody is in the DBA sysadmin group. How did you get there? It would fire off of one server in the domain but I don’t know anything about schemas.
Brent Ozar: Yeah, schemas are tricky because you can log DDL changes. The problem is if your trigger fails, then the change to the table can fail and that can be kind of ugly. You can also set up event notifications and dump stuff into a queue with Service Broker, but it is kind of challenging and kind of risky. If you want to learn more about it, search—god, I’ve got to type this woman’s name out—Maria Zakourdaev. So if you search for “event notifications and SQLblog,” that’s what you do: “SQLblog Maria.” SQLblog is all one word. Maria Zakourdaev, and I’m sure I’m butchering her name, from Israel has a post on how you go about setting up event notifications and how they break because they do break under some circumstances.
Erik Darling: Everyone mark it, not only with SQL Server 2016 release today but Brent recommended Service Broker.[Laughter]
Brent Ozar: It’s a great solution.
Richie Rump: You didn’t see the disdain on my face when he said that? You didn’t see that at all?
Jessica Connors: Let’s talk about snapshot replication from Trish L. “I have…
Tara Kizer: I’ve got to go get my coffee.
Brent Ozar: I know, we’re all like, “I’m out of here.”
Jessica Connors: Maybe we could tackle this. “I have snapshot replication which is scheduled to run one time per day but recently I’ve started to see blocking done by the snapshot replication. Do I need to [Inaudible 00:07:53] the distribution agent as well because it is running automatically now?”
Tara Kizer: I’m not sure about that but the blocking, you’re going to encounter that because it has to lock the schema. That’s one of the last steps it does. So anytime you have to do a snapshot, whether it be snapshot replication or transactional replication, I assume with merge replication too. Anytime you have to do that initial snapshot or reinitialize a snapshot, it does block changes—data changes, schema changes, you’ll see a lot of blocking as it’s going through the last bits of the snapshot creation.
Brent Ozar: What would make you choose snapshot replication? Like what would be a scenario where you’d go—or have there been any scenarios where you go, “Hey, snapshot replication is the right thing for something I encountered.”
Tara Kizer: I’ve never used to it but if users are willing to accept that their data is a day old, let’s say. Any time that I’ve used transactional replication, they’ve wanted near real time data. They wanted zero latency. We couldn’t deliver that in replication. But yeah, snapshot replication, it just depends on what your user wants as far as the data goes.
Richie Rump: I’ve used it for reporting solutions.[Richie and Erik speaking at the same time]
Jessica Connors: What?
Erik Darling: I was asking Tara if a different isolation level would help with that blocking.
Brent Ozar: Oh.
Tara Kizer: We were actually using RCSI so, yeah, it was definitely a schema lock. We definitely still had blocking.
Brent Ozar: Makes sense. It was probably worse without the schema, or without the snapshot or CSI, probably horrible.
Tara Kizer: It was very rare we had to do the snapshot but sometimes replication would be broken for whatever reason and we couldn’t figure it out and we’d just have to restart replication. Our database was large. It took like five to eight hours to do. Not the snapshot portion, the snapshot took like about 45 minutes I believe but there was a lot of blocking during that time.
Richie Rump: I like snapshot replication for reporting purposes, right? So again, you just dump the data over there and it’s okay that there’s a time delay for the reporting aspect and there’s your data.
Tara Kizer: I just wonder instead of snapshot replication if people should be, not backup and restore because that might take too long on larger databases, but a SAN snapshot, a daily SAN snapshot, because it’s just available right away. You don’t have to wait for anything.
Brent Ozar: No schema locks, it doesn’t matter what the volume of change data is, yeah.
Jessica Connors: While we’re on the hot topic of replication, there’s another one from Paul. “I am replicating a database using merge and had an issue where if the developers changed a procedure on the original database, the change would not be pushed to the replicated database. Replicate schema changes is set to true. Any guidance on the reason why the changes won’t replicate? I did a snapshot before initiating replication.”
Tara Kizer: So replicate schema changes has to do with the table changes, it does not have to do with stored procedure, views, functions, or anything like that. So if you do an alter table, add a column, that will get replicated if you have the replicate schema changes set to true but you would have to also have in a publication either your current publication or a different one to also replicate the stored procedures.
Brent Ozar: I wouldn’t do that in merge either. Like I would—if you’re going to change stored procedures, just keep them in source control and apply them to both servers.
Tara Kizer: Yeah.
Jessica Connors: Let’s move onto a question from Justin, SSIS Cache Connection Manager question. “I want to load several objects into cache, about one to five million records, but can’t figure out how to access that cache’s source of data. It’s quite a bit faster for us to load to a cache versus staging tables. Is this possible? If not, how would you store this?”
Brent Ozar: Have any of us used the caching stuff in SSIS? No, everybody is…
Tara Kizer: No, I’ve used SSIS a lot and have not used that.
Brent Ozar: The one guy I know who does is Andy Leonard. If you search for Andy Leonard SSIS, he’s written and talked about this. I know because it was in his book. I didn’t read the book, I just remember seeing the book. It was on my shelf at one time. It was a great paperweight. Smart guy, really friendly. Just go ask him the question, he’ll be able to give you that right away. Normally we’re all about, “Go put it on stack exchange.” Just go ask Andy. Just go “Andy Leonard SSIS” and he’s super friendly and will give you that answer right away.
Erik Darling: Tell him Brent sent you.
Brent Ozar: Tell him Brent sent you on this.
Jessica Connors: Question from Tim L. He says, “I’ve got an ancient Access expert here at my company. I’m having SA access. He has a lot of ODBC from multiple Access dbs into my 2008 R2 SQL Server. How do I find out what tables he updates? There’s nothing in terms of jobs or stored procedures that references his data pull and updates.”
Tara Kizer: You could do an Extended Event, run a trace, add a trigger.
Brent Ozar: It’s 2008 R2 though. I like the trigger.
Angie Rudduck: I like cutting his access.
Richie Rump: I love that, “ancient.”
Tara Kizer: Yeah, why does he need SA access?
Brent Ozar: Just go ask him. He’s ancient. He’ll be a nice guy. He’s mellow by now. If you run a trace, that’s going to be ugly, performance intensive. The trigger will be intensive.
Erik Darling: Well you can at least filter the trace down to table name.
Brent Ozar: Well but if he wants to know what tables he’s doing, it’s going to be every time…
Erik Darling: Oh, never mind.
Brent Ozar: Yeah.
Tara Kizer: He could filter by his login at least, if that’s what it’s going through at least to connect to SQL Server.
Brent Ozar: And don’t try to log his insert statements or updates deletes. Just put a record in a table the first time he does an update, delete, and then immediately turn off the trigger on that table, or the trace on that. But, yeah. That’s tough. Just go ask the guy. Go talk to the guy. It would be nice.
Erik Darling: Shoot him email.
Brent Ozar: Yeah, shoot him an email. Buy him a bottle of Bourbon.
Erik Darling: Yeah.
Brent Ozar: It’s a human being.
Richie Rump: Yeah, just don’t give away the wine. Right, Brent?
Brent Ozar: If you were going to give somebody wine, you should give them like Robert Mondavi.[Laughter]
Brent Ozar: He’s Access. He’s not, you know. That’s not true. Cliff Lede, ladies and gentlemen. This webcast is brought to you by Cliff Lede wines.
Jessica Connors: Do any of us participate in SQL Cruise?
Brent Ozar: I cofounded that with Tim Ford. Tim and I cofounded it and when we split up the consulting company versus the training and cruise-type business, I wanted to let him go do his own thing there and not be on it because I felt like I would kind of shadow in on it and make the thing murky. It is a wonderful experience. I strongly recommend it to anyone who thinks about going. It’s fantastic for your professional development. It’s limited to just say 20 attendees and like 5 to 10 presenters, so the mix, the ratio of presenters and attendees is fabulous. You get to hang out with them. You get to have dinners, from all of this you get to know them really well. So it can be a rocket ship for your career and it helps you really build networking bonds with not just the presenters but the other attendees who are there. The downside is you get to hang out with the presenters in hot tubs so that may be a pro or a con depending on what your idea of a good time is there. So it’s not for everybody but it is truly fantastic.
Erik Darling: Grant Fritchey in a speedo, ladies and gentlemen.[Laughter]
Jessica Connors: Do you still go on the cruise then? Are you done?
Brent Ozar: I don’t. I totally stopped doing that. I go off and do my own cruises. My next one is in Alaska in August I think, going on that one with my parents. But I haven’t done a technical cruise since. Most of the time what I like to do now is just go out on a cruise and not talk to anyone. I like to go out and sit and read books.
Erik Darling: You did Alaska before, right?
Brent Ozar: This is my fifth time I think, yeah. Absolutely love it. It’s gorgeous. I never was a snow kind of a guy but you get out there in the majestic snow and mountains and bears and all that, it’s beautiful.
Jessica Connors: Nice.
Angie Rudduck: Minus the jacket.
Brent Ozar: Yes.
Jessica Connors: Let’s talk to Graham Logan, he’s got some problems. He says, “SSMS crashes when expanding database objects in objects explorer. Database is about 1.2 terabytes and has about two million objects.”
Tara Kizer: Oh good lord.
Jessica Connors: But, he says, “[inaudible 00:15:43] design. It’s not mine. How to view all database objects without SSMS crashing.”
Tara Kizer: You just cannot use object explorer. You’re not going to be able to use object explorer. You can’t use the left pane in Management Studio. You’re going to have to write queries to see things. It’s very unfortunate but that’s a heck of a lot of objects in the database.
Brent Ozar: Before you expand the list, you have to right click on the tables thing and click filter. Then you can filter for specific strings but without filtering, it’s useless… I’d go information schema tables, information schema, yeah, all columns, all kinds of stuff.
Jessica Connors: Kyle Johnson has a new one. “We have a 4.2 terabyte database with a single data file. I’m working on a plan to migrate to multiple ones. Shrinking the database to a level with the data between files isn’t really practical with a six-hour window of no users. Have any other suggestions? Reindexing tables and specifying the file groups to move the table to two?” From Kyle Johnson.
Brent Ozar: Not a bunch of good options here.
Erik Darling: Brent is getting ready to tell you about Bob Pusateri.
Brent Ozar: I was. You are psychic. You are phenomenally psychic. Tell us more. I want to subscribe to your newsletter.
Erik Darling: Bob Pusateri, which I feel like this webcast has been obscene enough without me saying that, has a blog post about moving file groups, a lot of the gotchas, and you know, bad things that can happen to you. I will track down the link for it and send it to you but I would not do it justice just explaining what goes on it, because it’s scripts and everything, so.
Brent Ozar: Bob had a 25 terabyte data warehouse with thousands of files in it because the prior DBA thought it was a good idea to create a separate file group for every employee and then later came to regret that decision so he has a great set of scripts on how you go about moving stuff around and keeping them online wherever possible. So it’s really slick. So you do that prepping leading up to the six-hour window so that your six-hour window is only dealing with stuff that you can’t do offline, like moving the LOB data if I remember right.
Jessica Connors: Question from Claudio. “I’m trying to understand the differences between the new AlwaysOn Basic Availability Groups and the synchronous commit mode and the mirroring and high safety mode but they look identical except AlwaysOn seems more complicated to set up and manage. Are there any benefits to either solutions, features, performance, licenses, liability? Which one would you recommend to adopt?
Tara Kizer: Database mirroring is being deprecated so you’re going to want to move over to the AG basic availability groups. Get on it now. It’s the replacement for database mirroring.
Brent Ozar: The drawbacks, so you’ve managed both too. What would you say the strengths of AlwaysOn Availability Groups are over mirroring and vice versa? That’s not a trick question, I promise.
Tara Kizer: Mirroring you’re not failing over groups at a time. You’re failing over a database at a time. So availability groups let you failover in groups which is good when you have an application with multiple databases that it needs.
Brent Ozar: To be clear, so you’re saying the guy is saying Standard too, so you only do one database at a time. You could script those too, just like you would with mirroring. I’m trying to think if there’s anything that would be… have to have a cluster but you don’t have to have a domain with mirroring. But you don’t either with 2016 either. You can do it between standalone boxes.
Tara Kizer: With mirroring, if you want the automatic failovers, you need a witness. With AGs you do need a quorum but it could be a file share on another server, you know, on a file server that you have or a disk on a SAN could be a quorum. Mirroring does require another box, a VM, it can be Express Edition.
Brent Ozar: Yeah, I used to be the biggest fan of mirroring. I’m having a tough time coming up with advantages as 2016 is starting here.
Tara Kizer: I did a lot of failovers with mirroring, log shipping, and then later availability groups and by far I like availability groups best for DR failovers. It was just so much easier. You just run a failover command and you’re done. With mirroring, you’re doing it database by database. Log shipping is, you know, all sorts of restores going on. Mirroring is certainly easy, definitely easy, but I like the slickness of availability groups and readable secondaries and the choice of asynchronous and synchronous.
Brent Ozar: Yeah, that’s where I was going to go too. Because even in Standard, you get choice between synch and asynch now. And you can use one technology that works on your Standard stuff and your Enterprise stuff so you only have to learn one feature instead of learning two. That’s kind of slick too.
Tara Kizer: When we used mirroring, we would use asynchronous mirroring to the DR site then for high availability solution at the primary site we used failover clustering. So availability groups it just solves both solutions in one feature, plus reporting, we got rid of replication.
Jessica Connors: All right. Let’s move on to a question from Chris Woods, a regular attendee. He says, “Migrating MDF with LOB data, L-O-B data, I don’t know how you call that, from one drive to another with minimal/no downtime. Can you use log mirroring to mirror it to a new database on the same server that shut down the original during a quick downtime?”
Brent Ozar: You can’t do mirroring to the different database on the same server, can you? You can do log shipping, can you do mirroring to the same instance?
Tara Kizer: No.
Brent Ozar: You can do log shipping to the same instance. That will work. Your downtime will be super fast. Because what your failover process would look like is when it comes time for failover, you do a tail of the log backup up on the main database, then restore that tail-log on the other database. Rename the old primary as like, the old primary database just like “database old.” Then rename the new one as “database new” and then whatever the new database name is or the original database name is. So you could do that in like a 30-second outage. You don’t have to change connection strings because it’s all the same server still. So that’s kind of slick.
Tara Kizer: If this is a SAN drive, even moving from one SAN to the next, we did all this stuff live. I don’t know what the technologies are called but we would move arrays live. The SAN administrators did some magic and it just copied over the data and once the copy was complete, it did a switcheroo between the two pointers, or, I don’t know what the technology was but the SAN can handle this without any down time.
Jessica Connors: Rob is adding a new instance to an existing active active cluster. I think he’s talking me about his process so that we can say if it’s yea or nay. He says, “I would need to failover the existing instances to one node, install the new instance on the node with no instances. Service pack it up and failover the instance to the node I was just on. Then run the install in another node, apply SPS, then rebalance the instances.” Does that sound about right?
Tara Kizer: It does but you know we don’t recommend active active clusters. What happens if you lose a node? I don’t at least. I’ve had four-node clusters where all four nodes were active. It’s just a nightmare. If you lose a node, can your other nodes support all of the instances at the same time until you get that other node fixed?
Brent Ozar: Richie is showing something on his iPad. What I would say is…
Erik Darling: It’s too bright.
Brent Ozar: I still can’t see it. We do recommend active active with a passive node.
Tara Kizer: Yeah, okay. Right.
Brent Ozar: Yeah, multi-instance clusters, just have a passive in there somewhere. Your scenario is exactly why you want a passive node laying around.
Tara Kizer: At least what you wrote out here for the question, yeah, that is the process.
Brent Ozar: Also known as miserable.
Tara Kizer: Yeah. At least since SQL Server 2008 we’ve been able to have where it can install it just on one node. Prior to that, all nodes in the cluster had to be online and have the exact right status in order for the installation. Because the installation occurred across all nodes at the same time. Service packs, the engine, everything. On a four-node cluster, there’d always be one node that was misbehaving. It just says, “I need a reboot.” And you’d reboot it 20 times and it would still say, “I need a reboot.” Then finally that one would be okay and now another node would say, “I need a reboot.” It was just ridiculous. So I’m glad that Microsoft changed the installation process starting with 2008.
Brent Ozar: It’s like taking kids on a road trip. “Everybody ready…?” “No.”
Erik Darling: “I have to pee.”
Richie Rump: I got excited, I thought we had a Node.js question but I guess not.
Erik Darling: Never have, never will.
Richie Rump: Brent has.
Brent Ozar: I have.
Jessica Connors: Let’s take one more question. Let’s see here. “Good morning, Brent and Tara, Erik, Richie, and Angie. Says, “Yesterday we had a problem with the process that normally moves data from table queue and delete it after it’s done. This is a standalone database. We stopped the inflow of data but it didn’t help. We got thousands of deadlock alerts. I notice that the disk queue length on the log drive is higher than usual. Here is a sample of the deadlock.” He provides it. “Is there anywhere I could look for this issue?”
Tara Kizer: If you’re getting deadlocks you should turn on the deadlock trace flag 1222, maybe run an Extended Event to capture the deadlock graph. Having just the deadlock victim isn’t enough to be able to resolve it.
Brent Ozar: It’s a separate technique I think that not a lot of database administrators get good at because it’s one of those things where you’re kind of like, “Hey, you should fix your indexes in your queries.” Then people go off and do their own thing. It’s one of those where when you do want to do it, it takes a day or two to read up and go, “Here’s exactly how the [Inaudible 00:25:00].” There’s also not a lot of good resources on our site for it. We don’t go into details on deadlocks either. Have any of you guys seen resources on deadlocks that you liked?
Erik Darling: I like just hitting Extended Events for it. The system health session has quite a bevy of information on deadlocks and you can view the graphs and everything which is pretty swell.
Tara Kizer: I attended a session at PASS in 2014, Jonathan Kehayias from SQLskills, it was all about deadlocks. It was invaluable information. He went over different scenarios and stuff. He said that he loves deadlocks. It was like, whoa, I don’t know that anyone has ever said that before. But it was really great information. I haven’t looked at—I do read his blogs—but I suspect he’s got a lot of deadlock information on the blog to help you out.
Richie Rump: He also loves XML.
Brent Ozar: He loves XML and Extended Events. If you have a Pluralsight subscription. So Pluralsight has online training. I want to say it’s like $39 a month or something like that. I think Kehayias has a course on deadlocks. I’m not 100 percent sure but if you search for SQL Server deadlocks if Kehayias has a course on there, it would be wonderful.
Erik Darling: Also, if you don’t have Pluralsight but you want to try it, Microsoft has a Dev Essentials site I believe where if you sign up for that, you get a 30-day free trial of Pluralsight and you also get Developer Edition and a copy of Visual Studio that’s free, Visual Studio Community or something for free. So it’s not just the Pluralsight subscription for 30-days but you do get a couple other goodies in there that last you a little bit longer.
Richie Rump: The course is called SQL Server Deadlock Analysis and Prevention.
Angie Rudduck: Someday still has a Pluralsight account.
Jessica Connors: All right guys, that’s all we’ve got for today.
Brent Ozar: But thanks for hanging out with us. Man, time goes so fast now. Gee, holy smokes.
Erik Darling: And they’re sobering up.
Brent Ozar: Well, back to work. The Cliff Lede, ladies and gentlemen. Enjoy the High Fidelity. See you guys next week.