July 2011 - SQLHA

July 25, 2011

Publishing: Physical or Not?

As I plow ahead with my plans for the Denali version of my book (good suprises to be announced when I talk about it in detail soon), I’m doing my research. In not wanting to compromise the content (meaning a larger page count), I keep running into one constant roadblock: a physical book and its cost to both you and me. This challenge has a few aspects.

First and foremost, let me say that I love “real” books and in tech books, I often prefer them. However, the higher the page count, the more the cost goes up. It’s a conundrum for authors such as myself that want to tackle topics properly. No one (including myself) is going to get rich off of doing a book, so that isn’t my motivation. If you want an eBook, that’s not going to be a problem. I own an eBook reader (the Sony PRS-950, and the PRS-505 before it). I do like eBooks as well, and the new book – no matter what the page count – will be available in the most popular formats (currently targeting PDF, ePub, and Kindle).

The publishing world has changed. With the closing of Borders here in the USA, my chances of getting the book on a shelf decreases immensely. Sure, we’ve got Barnes & Noble and smaller independents, but they’ve also got limited (and shrinking) shelf space that probably won’t be taken up by a self-published book even if they can get it from Ingram. I’m realistic. So why would I go through all the trouble to sell very few physical copies?
The number of print on demand publishers that can handle a book of the size I’m proposing is limited. I’m looking into costs now of what this will be. I really don’t want to charge you guys an arm and a leg just to get a physical book.
According to some research I’ve done, Amazon is not keeping print on demand books that they don’t publish. If I go with someone other than Create Space, that makes getting a physical book frustrating for everyone since Amazon is a pretty big distribution channel. Sure, that means you will be able to order it online, but from where? Obviously I can make it available via my site and I’m hoping it would show up elsewhere, but this is a very real and valid concern of mine.

I think many of you would like a physical option and I’m really trying to make that happen and at a reasonable cost. Will it be possible? Time will tell. Here are some options as I see them:

Print a limited run of physical books that would be done via preorders only. Once they’re gone, they’re gone and eBook will be the only option. Maybe I would do another run if demand was there to do another run. I live in an apartment and really don’t have the space to have cases and cases of books that will never sell. That’s just an expense and physical limitation I cannot afford.
Figure out the whole distribution thing with print on demand, but that is still a really big question mark. This would allow you to get a printed book until the end of time. This would clearly be the best option if it’s available. Stay tuned.
Only offer an eBook, and eschew the idea of a physical book altogether. This would sadden me – trust me. This is a last resort option. However, it would allow you to print the sections you want on your own. For example, want only the clustering how-to stuff printed? That’s all you would get if you printed that yourself. I see a big upside to choice in doing that. If I do go down this road, I would probably offer some sort of multi-format deal allowing you a copy for your eReader and the straight PDF, which would make it easier to print. Obviously in terms of cost, having just an eBook will keep things reasonable, even with a larger reference book as the one I am aiming to do.

There’s one other distinct advantage to only being an eBook: I can publish very soon after Denali is released. That’s a pretty big advantage to everyone – readers and author alike!

Note to publishers out there: if you’re willing and interested to take this book on, I’m all ears. Use the Contact Allan link above.

July 25, 2011

Un-SQL Friday #005: Bad Interviews

Jen McCown (blog | Twitter), one half of DBAs at Midnight, put out the request for bad interview stories, so I figured I’d chime in. I’ve got two stories.

As a Job Candidate

I think this happened in the late 90s, during the process which led up to me getting my job with a small corporation named Microsoft. Maybe you’ve heard of them. A little context: this was in the February/March timeframe here in Massachusetts, so it was still somewhat cold with snow still on the ground in parts. I knew it was time to look for a new job, and both of my bosses were out for about two weeks so I had a golden opportunity to make the most of it.

I was working with a recruiter or two. First of all, let me apologize to the good recruiters out there. Unfortunately for their profession, I’ve found there are very few good ones out there. I’ve likened them to car salesmen. To this day I get pings from people who got my resume years ago (haven’t submitted for a job in probably 7 years going back to Avanade) and want updated info such as what I’m doing, etc. Recruiters could be the subject of a whole post, but I won’t go there. They do serve a means to an end (better to have more eyes out there looking assuming they get what you want to do).

When it comes to working with recruiters, often times I’ll do the “gimme” interview. A “gimme” interview is one you know probably won’t wind up being a job for you, but shows you are willing to work with the recruiter. Sometimes you just have to do it, although these days I’m not sure I’d waste my time. Anyway …

This company was out I think in Concord or something like that. My interview was at 9AM, and proper interview etiquette is to get there at least 15 minutes early. I left around 8 and was there by 8:30. Anyone who has driven Route 128 (nee 95) knows depending on time of the day and weather, it can get backed up in spots. I remember one time when it started snowing while I was an intern at Sybase in Burlington, MA, it took me 3 hours to get from Burlington to Waltham back to my dorm. That’s a 20 minute or so drive not in traffic to put it in perspective.

I sat in my car for a few minutes and walked up to the door. Lights were off, everything was locked. Couldn’t get in. There wasn’t even a receptionist. This is never a good sign. I’m not a morning person (just ask anyone who knows me), but no one at a company before 9AM? Hmm. I believe the receptionist showed up around 8:50, and was surpised to see me there. I took a seat in the reception area.

Finally around 9:15 or so someone comes to get me and take me into a conference room. They give me the standard application form to start filling out (which I should have been handed earlier, but whatever). That person doesn’t come back for at least 15 minutes. It may have been longer. Needless to say, my patience is starting to wear thin and I haven’t really talked to anyone yet. <rant> Here’s a lesson out there for future employers of me or anyone else you are trying to woo: you guys are selling us on you as well. It’s not just the interview candidate trying to sell themselves. The impression I was getting was one of disorganization and not caring. When you set up an interview loop, make sure everyone is aware of it and it runs smoothly as well as relatively on time. </rant>

As if you couldn ‘t tell already, things were going downhill faster than a speeding bullet. I had already pretty much made up my mind this company wasn’t going to be for me, but I would see how things went. Worst case, I would have fun with the interview. And I did. I don’t recommend some of the behavior I am going to talk about unless you really want to sabotage an interview.

The first person comes in to interview me and it was clear they had just gotten the tap on the shoulder to come talk to me. I know that happens (I’ve been on the other end of that where I had to interview someone on short notice), but at least come somewhat prepared. So interview #1 was not awe inspiring and I played nice. I didn’t get a good feel of anything – what the job I was going to do would be, the environment, etc. I then wait at least another 30 minutes for interview #2, and I am seething at this point. I’ve spent more time waiting than anything else, so I decide to torpedo things and end the pain. I think I answered every question with “Yes”, “No”, or some really short answer that had no way of follow up. That second interviewer also rubbed me the wrong way with a bit of attitude, so I gave it back. I was very confrontational and a bit aggressive. You may think that’s rude or wrong, but I knew I wasn’t going to work there, and I definitely knew that I had been there close to 2.5 – 3 hours with not much done. News at 11: I was told shortly thereafter they would get back to me and I could go.

I can remember calling the recruiter and unloading on them about the experience and they were trying to convince me that the experience was not the norm. Forget that. I never worked with that recruiter again. It’s also why I will never do the “gimme” again. I’ve always had a good gut instinct that rarely fails. In the job hunt process sometimes you need to kiss a few frogs before finding your prince or princess (whatever your preference may be).

As an Interviewer
This is one interview I’ll never forget. I was working for a consulting firm, and the way they do interviews is bring a bunch of people around in a single day. I was asked to be part of the interview loop for SQL folks that made it through the phone screen. Whenever I do interviews, I don’t regurgitate their resume. In a consulting role, I’m trying to assess a few things:

Do they have a good skillset?
Can they interface with customers/have good people skills?
Could I work with this person onsite and trust them?

While similar to a normal 9-to-5, a consultant needs to have a few other things that you can sometimes hide or not care about if you’re in cubeland 40 hours a week.

Anyway, whether in person or on a phone screen, I also don’t ask things like “What’s a clustered index?” unless I feel they gave me some answer and they have no idea what they’re talking about. My job isn’t to figure out if you can remember the Nth clause of a WITH statement in some T-SQL command. That can be taught or looked up. Good problem solving skills and troubleshooting methodology is hard to come by.

I’m in my early 30s at the time (I’ll be 40 later this year), so I didn’t have all the experience I have now, but I knew what we needed. This guy was in his 40s, and was supposedly some seasoned veteran. I tend to ask situational type questions that will force candidates to give me a combined technical as well as non-technical answer (for example, how to deal with the problem and the people involved). His first answer troubled me as to his “expertise”, so I gave him a few more and wasn’t convinced that this guy was the expert he thought he was. I asked him a few just technical questions and I was not getting a good vibe either. So I ended the interview and gave my feedback.

The guy was not hired, and I heard that he complained to HR that he didn’t feel he got a fair interview because the people who talked to him (including myself) didn’t ask tough enough questions or something like that. Unbelievable, right? Chutzpah is the only word I can use here to describe that move. To a person, every since one of us that talked to him got the same impression. Just because you’ve been in the industry for years doesn’t mean anything. You could have 10 years of experience but done the same thing for 10 straight years with no growth. It doesn’t mean you’re bad, it means that compared to everyone else out there, even with some with less years of experience, you may still be considered junior. That’s reality. Try to do different things even if you are at the same employer. It only makes you more marketable longer term even if you have no intention on leaving your job.

So to sum up here: never oversell yourself as someone you’re not. Interviewers will figure you out. It’s one thing to be confident, it’s another thing to lie or be wrong and then have the cojones to complain about the interview later. If you don’t know, say, “I don’t know.” I respect that much more.

July 21, 2011

Do You Like Your Writing Boring? Me Neither

In a brief Twitter conversation with Wendy Pastrick (blog | Twitter), who is also another great SQL person out there (just incase you didn’t know), she had this to say:
“I do wish whitepapers were more interesting of a read at times. I’ve fallen asleep on a few.”
I couldn’t agree more, and let me expound on that. (Hey, it’s my blog so I could do it anyway!)

If you don’t know, I’ve written quite a bit over the years (check out some of the links on the main SQLHA.com site which has some of them) including the books I’ve authored or co-authored. I’m always doing some kind of content, and doubt I’ll ever stop. However, when I’ve done some writing for clients, every now and then I’ll get the comments back that the writing style is not formal enough or that I use some colloquial phrases which won’t fly (oops! there’s one …). While I see their point, and in some cases they are probably right, I’ll go back to Wendy’s quote. Most people would rather it be a bit less formal and easy to read than stiff as a board. There is a happy medium somewhere that blends readability with great technical content. They do not need to be mutually exclusive.

I’m not saying more formal writing does not have its place nor were my reviewers wrong. In some cases over the years, reviewers were absolutely, 100% correct. Sometimes things can be a little too informal … and that’s why I have reviewers actually look at not only accuracy of content, but readability. Both are very important not only to me, but you, the reader. If you have content that is well written but sucks in terms of its technical accuracy, why even bother? Same for the reverse: if it is technically right but hard to read and ultimately misses its audience, it’s a waste of time for everyone involved.

I do want to say one thing about reviewers. I always read every comment they write – good or bad. Good ones will call me out on everything, even down to a missing period or a misspelling. If I blundered something technical, absolutely flog me. I have no ego about it. Did I completely miss something I should have talked about? Let me know. Sometimes it’s possible to add, in other cases, no for various reasons that could be their own blog post (ah, politics!).

With lots of reviews come lots of opinions. I often get polar opposites (loved it/hated it). At the end of the day, I need to figure out which one to go with and stick to my decision. I try to take consensus. If there is a lone dissenter, they’re often an outlayer (outside of being technically wrong). If it’s a lot of folks, genuine issue. If it’s some but not all, I need to think about it a bit more. One of the advantages of having multiple reviewers is that it’s invaluable, but it’s also a bit of a curse because it does add a lot of work for an author.

Back to the topic at hand. If doing a whitepaper or training course for someone like Microsoft or another company (which may or may not have my name attached to it), chances are it needs to be a bit more formal. Using colloquial phrases can be problematic even in informal writing because they are not always universal around the world. For example, saying something like, “Boy, implemeting things like this is like taking a swing for the fences and striking out.” While you can get the meaning from the context, it’s a baseball reference. Baseball is far from a universally known sport, let alone its idiosyncrasies and nuances. Over the years I’ve become much better at recognizing that and most of it gets excised before final drafts, but it does not always happen. So I apologize in advance if you do not always get some of that stuff in a publication that has me associated with it.

Having said that, I do rely on editors in this entire process. While I try to catch a lot of this kind of stuff, it’s really up to them to a large degree to try to fix any wording or verbiage that is not up to snuff, or mark it so I can deal with it. Good editors excel at this and bad ones miss it entirely. Every client may even have a “house style” as to how they prefer something to sound/read, and I’m fine with that – they just need to let me know.

Bottom line: realize before you get something to read, it’s been through a lot of people’s hands. One thing with editors that I’ve found is that if they are unfamiliar with SQL Server, sometimes they’ll alter wording in a way that is wrong so I’ll have to work with them to come up with something that works for both of us. Trust me, there’s a lot of work that goes into writing a document that no one ever sees. If it was easier, I think more people would do it (if they like to write, that is; most people hate to compose documentation of any sort).

I’ve had quite a few people over the years after reading some of my stuff (especially my 2005 and 2008 books) that they can hear “my voice” in them. That’s great, because I tend to think that people who know me or have possibly seen me speak can see that. One of the hardest things to do is take concepts that could be difficult or complex, and make them either relatable or easy to understand. For me, it’s not unlike writing a fiction story. What do I mean by that?

Whenever I sit down to write something, especially a whitepaper or book, I try to figure out who my audience is and what story I want to tell. You need to start out at A and end up at Z. The thoughts need to be coherent and build on one another. For example, I’ve always maintained you can’t really talk about SQL Server failover clustering without understanding and explaining the Windows side of things. It’s why I put it in my books and cover it before I talk about the SQL part. However, when I explain the Windows side, I do it from a SQL perspective and address it that way so it ties everything together. So even though Windows is a bit of a left turn, it’s all part of the same thing. It’s up to me as the author, to tie it all together and wrap it up into a neat bow.

A big thanks to Wendy for inspiring this blog post, and all of this is quite relevant as I embark on the planning process for the Denali followup to the 2000, 2005, and 2008 books. By the time you see that book, the content will have been vetted technically by reviewers, massaged and altered (probably a few times), and had its makeup applied so it looks pretty for you. I’ll make a more formal announcement of my plans for the Denali book soon … it’s shaping up nicely.

July 20, 2011

SQL Server Denali CTP3: Skipping Update Checks via Command Line Installation

If you have installed SQL Server Denali CTP3, whether on a cluster or standalone, you may have noticed this brand new dialog during installation:

Figure 1. Update Check in Denali Setup

The skip GUI does not work very well to be honest with you. I find it still tries and ultimately fails even if you tell it to skip the check or unselect Include SQL Server product updates. Because of that, the GUI right now for me is a longer process than it needs to be. I prefer to do command line installations overall, but do GUI as well. Especially since Denali is CTP, I want to test as many code paths as possible. What can I say? I never got rid of the old QA engineer/tester in me.

If you are doing command line or file-based scripted installs, there is good news. There is a switch/command for SQL Server Denali that will allow you to completely skip the update check and it does work. The switch is UPDATEENABLED. Set it to false and Setup will not check for updates online. This is especially helpful for installations that will not have Internet access.

The syntax would look like this:
Command Line (as part of a bigger string of commands)
[cc]/UPDATEENABLED=”FALSE”[/cc]

In a file (on its own line)
[cc]UPDATEENABLED=”FALSE”[/cc]

In the log file, you will see this:
[cc width=”700″]
(04) 2011-07-19 20:29:17 Slp: ———————————————————————-
(04) 2011-07-19 20:29:17 Slp: Running Action: SearchUpdatesAction
(04) 2011-07-19 20:29:17 Slp: Skipping update search
(04) 2011-07-19 20:29:17 Slp: Completed Action: SearchUpdatesAction, returned True
(04) 2011-07-19 20:29:17 Slp: ———————————————————————-
[/cc]

Of course, if you want to check for updates, don’t add this line which means the default value of TRUE is set.

Have fun with command line installations!

July 12, 2011

SQL Server Denali CTP3 Clustered Instance Installations, Backslashes, and You

If you haven’t heard by now, SQL Server Denali CTP3 (build 11.0.1440.19 if you are keeping track) is available for download on both 32- and 64-bit flavors. While I still don’t agree with a 32-bit version being released, it’s nice to be able to talk about something other than CTP1. I haven’t talked about Denali much because I’ve been under NDA with some private builds where I was lucky enough to preview some things, and with CTP3, I can finally talk about them – especially with regards to AlwaysOn availability groups. I won’t be addressing them in this first Denali-related post, though. I’ll be talking about something near and dear to my heart: failover clustering.

The installation of a clustered instance is similar to that of SQL Server 2008 and SQL Server 2008 R2. Things get a bit different on the Add Node operation, but that will be saved for another day.

SQL Server 2008 introduced the ability to create the resource group in the Windows failover cluster that will contain the clustered SQL Server resources. As long as the character was valid in the Windows failover cluster, you could also use it successfully in the SQL Server 2008 and SQL Server 2008 R2 installations with no problems. For example, when configuring a named instance which is really NETWORK_NAMENAMED_INSTANCE_PART, I would always name the group the same as the instance so it’s easy to identify in Failover Cluster Manager (or PowerShell in Windows Server 2008 R2). So if I install an instance with a network name of GRAND and a named instance of ILLUSION, I would use GRANDILLUSION. A sample is shown below in Figure 1.

Figure 1. Creating a resource group for use with SQL Server

However, a note of caution in SQL Server Denali’s installation related to this screen that could trip you up: do not use a backslash (). It will cause Setup to ultimately fail. It won’t be immediate. It will be most of the way through the install (unfortunately). When you get to the step cluster_group_cluster_config_Cpu64. You will see the error in Figure 2 if you are using the Setup user interface. Ironically as you can see, it creates the group just fine but can’t do anything else.

Figure 2. Setup failing when a slash is used.

This is what you will see in the log:
(01) 2011-07-12 15:28:39 Slp: SetValue: HKEY_LOCAL_MACHINESOFTWAREMicrosoftMicrosoft SQL Server110ConfigurationState, Name = SQL_PowerShell_Engine_CNS
(01) 2011-07-12 15:28:39 Slp: Action “ConfigEvent_SQL_Engine_Core_Inst_sql_engine_core_inst_Cpu64_Install_Finalize_finalize” will return false due to the following conditions:
(01) 2011-07-12 15:28:39 Slp: Condition “Feature dependency condition for action: ConfigEvent_SQL_Engine_Core_Inst_sql_engine_core_inst_Cpu64_Install_Finalize_finalize The condition tests feature: SQL_Engine_Core_Inst_sql_engine_core_inst_Cpu64. There are 10 dependant features. The feature is tested for results: ValidateResult, Result, CleanupResult.” did not pass as it returned false and true was expected.
(01) 2011-07-12 15:28:39 Slp: Condition is false because the required feature SQL_Engine_Core_Inst_sql_engine_core_inst_Cpu64 failed in result Result
(01) 2011-07-12 15:28:39 Slp: Action “cluster_group_cluster_finalconfig_Cpu64” will return false due to the following conditions:
(01) 2011-07-12 15:28:39 Slp: Condition “Feature dependency condition for action: cluster_group_cluster_finalconfig_Cpu64 The condition tests feature: SQL_Engine_Core_Inst_sql_engine_core_inst_Cpu64. There are 10 dependant features. The feature is tested for results: ValidateResult, Result, CleanupResult.” did not pass as it returned false and true was expected.
(01) 2011-07-12 15:28:39 Slp: Condition is false because the required feature SQL_Engine_Core_Inst_sql_engine_core_inst_Cpu64 failed in result Result
(01) 2011-07-12 15:28:39 Slp: Action “ConfigEvent_MPT_AGENT_CORE_CNI_sql_engine_core_inst_Cpu64_Install_Finalize_finalize” will return false due to the following conditions:
(01) 2011-07-12 15:28:39 Slp: Condition “Feature dependency condition for action: ConfigEvent_MPT_AGENT_CORE_CNI_sql_engine_core_inst_Cpu64_Install_Finalize_finalize The condition tests feature: MPT_AGENT_CORE_CNI_sql_engine_core_inst_Cpu64. There are 2 dependant features. The feature is tested for results: ValidateResult, Result, CleanupResult.” did not pass as it returned false and true was expected.
(01) 2011-07-12 15:28:39 Slp: Condition is false because the required feature SQL_Engine_Core_Inst_sql_engine_core_inst_Cpu64 failed in result Result
(01) 2011-07-12 15:28:39 Slp: Action “ConfigEvent_SQL_Replication_Core_Inst_sql_engine_core_inst_Cpu64_Install_Finalize_finalize” will return false due to the following conditions:
(01) 2011-07-12 15:28:39 Slp: Condition “Feature dependency condition for action: ConfigEvent_SQL_Replication_Core_Inst_sql_engine_core_inst_Cpu64_Install_Finalize_finalize The condition tests feature: SQL_Replication_Core_Inst_sql_engine_core_inst_Cpu64. There are 4 dependant features. The feature is tested for results: ValidateResult, Result, CleanupResult.” did not pass as it returned false and true was expected.
(01) 2011-07-12 15:28:39 Slp: Condition is false because the required feature SQL_Engine_Core_Inst_sql_engine_core_inst_Cpu64 failed in result Result
(01) 2011-07-12 15:28:39 Slp: Action “ConfigEvent_SQL_FullText_Adv_sql_fulltext_Cpu64_Install_Finalize_finalize” will return false due to the following conditions:
(01) 2011-07-12 15:28:39 Slp: Condition “Feature dependency condition for action: ConfigEvent_SQL_FullText_Adv_sql_fulltext_Cpu64_Install_Finalize_finalize The condition tests feature: SQL_FullText_Adv_sql_fulltext_Cpu64. There are 3 dependant features. The feature is tested for results: ValidateResult, Result, CleanupResult.” did not pass as it returned false and true was expected.
(01) 2011-07-12 15:28:39 Slp: Condition is false because the required feature SQL_Engine_Core_Inst_sql_engine_core_inst_Cpu64 failed in result Result
…
…
(01) 2011-07-12 15:29:10 Slp: Error result: -2032664561
(01) 2011-07-12 15:29:10 Slp: Result facility code: 1752
(01) 2011-07-12 15:29:10 Slp: Result error code: 15
(01) 2011-07-12 15:29:11 Slp: WER: Application level consent value ‘False’ was mapped to consent status ‘WerConsentDenied’
(01) 2011-07-12 15:29:12 Slp: WER: Result of the submission:: ‘WerDisabled’
(01) 2011-07-12 15:29:12 Slp: WER: Submitted 1 of 1 failures to the Watson data repository

BTW, I love the spelling of dependent (see bold). Although I’ll take bad spelling here since the verbose log is pretty good.

Anyway, the next time you go to run Setup, here’s what you will see on the instance screen as shown in Figure 3:

Instance Configuration dialog after the failure

Figure 3. Instance Configuration dialog after the Setup failure

What this means is simple: SQL Server Setup failed and didn’t clean up after itself as evidenced by the part I put in a red box. It thinks that the MOVINGPICTURES instance was installed. You can verify this by trying to install that instance again if you’re a masochist. You can see what that error looks like in Figure 4.

: Figure 4. Trying to reinstall MOVINGPICTURES

At this point, the logical thing is to run the uninstall procedure (remove a node) to get rid of the failed installation. Note the blank Network Name as shown in Figure 5.

Figure 5. Uninstalling the failed instance

This seems to work fine, but if you then go to install the instance and do not reboot, you will see the error in Figure 6 after it installs the Setup files. Uh oh! This one I need to file a bug on.

Figure 6. Error after uninstall and attempting reinstall

So I reboot to clear the proverbial decks. In looking around after the reboot, Setup does not clean up the MOVINGPICTURES group it created in the Windows failover cluster. You have to do that manually. Now, I’m not sure if it was the reboot or the deletion of the resource group, but you can install the instance if you re-run the first node installation as shown in Figure 7. Note that the shared files and utilities which I selected as part of the original install are still there. SQL Server only installs those files once, so at this point, it will just install the components for the instance. This is true of SQL Server 2008 and SQL Server 2008 R2 as well.

Figure 7. Clean Instance Configuration dialog after uninstall

When done, you’ll wind up with an instance in a group with no backslash as shown in Figure 8.

Figure 8. MOVINGPICTURES finally installed

You then decide since it’s 100% legal in Windows failover clustering to rename the group to be MOVINGPICTURES as shown in Figure 9. You should have no problems going forward, right? Wrong.

Figure 9. Adding the slash back into the resource group

For example, If you try to install another instance of SQL Server things will seem fine for a bit. When you click Next on the Disk Space Requirements dialog, and it attempts to bring up the one for teh resource group, you will now fail with the error shown in Figure 10.

Figure 10. Another first node setup fails due to a slash ... but it's not obvious

It’s a cryptic message, but it stems from renaming the resource group and adding the backslash. Here’s what you see in the log:

(01) 2011-07-12 17:16:01 Slp: Error: Action “Microsoft.SqlServer.Configuration.UIExtension.WaypointAction” threw an exception during execution.
(01) 2011-07-12 17:16:01 Slp: Microsoft.SqlServer.Setup.Chainer.Workflow.ActionExecutionException: Thread was being aborted. —> System.Threading.ThreadAbortException: Thread was being aborted.
(01) 2011-07-12 17:16:01 Slp:    at System.Threading.WaitHandle.WaitOneNative(SafeWaitHandle waitHandle, UInt32 millisecondsTimeout, Boolean hasThreadAffinity, Boolean exitContext)
(01) 2011-07-12 17:16:01 Slp:    at System.Threading.WaitHandle.WaitOne(Int64 timeout, Boolean exitContext)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Configuration.UIExtension.Request.Wait()
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Configuration.UIExtension.UserInterfaceProxy.NavigateToWaypoint(String moniker)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Configuration.UIExtension.WaypointAction.ExecuteAction(String actionId)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Chainer.Infrastructure.Action.Execute(String actionId, TextWriter errorStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionHelper(TextWriter statusStream, ISequencedAction actionToRun)
(01) 2011-07-12 17:16:01 Slp:    — End of inner exception stack trace —
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionHelper(TextWriter statusStream, ISequencedAction actionToRun)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionWithRetryHelper(WorkflowObject metaDb, ActionKey action, ActionMetadata actionMetadata, TextWriter statusStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.InvokeAction(WorkflowObject metabase, TextWriter statusStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.PendingActions.InvokeActions(WorkflowObject metaDb, TextWriter loggingStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionEngine.RunActionQueue()
(01) 2011-07-12 17:16:01 Slp: Error: Action “Microsoft.SqlServer.Configuration.BootstrapExtension.ExecuteWorkflowAction” threw an exception during execution.
(01) 2011-07-12 17:16:01 Slp: Microsoft.SqlServer.Setup.Chainer.Workflow.ActionExecutionException: Thread was being aborted. —> System.Threading.ThreadAbortException: Thread was being aborted.
(01) 2011-07-12 17:16:01 Slp:    at System.Threading.WaitHandle.WaitOneNative(SafeWaitHandle waitHandle, UInt32 millisecondsTimeout, Boolean hasThreadAffinity, Boolean exitContext)
(01) 2011-07-12 17:16:01 Slp:    at System.Threading.WaitHandle.WaitOne(Int64 timeout, Boolean exitContext)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Configuration.UIExtension.Request.Wait()
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Configuration.UIExtension.UserInterfaceProxy.NavigateToWaypoint(String moniker)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Configuration.UIExtension.WaypointAction.ExecuteAction(String actionId)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Chainer.Infrastructure.Action.Execute(String actionId, TextWriter errorStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionHelper(TextWriter statusStream, ISequencedAction actionToRun)
(01) 2011-07-12 17:16:01 Slp:    — End of inner exception stack trace —
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionHelper(TextWriter statusStream, ISequencedAction actionToRun)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionWithRetryHelper(WorkflowObject metaDb, ActionKey action, ActionMetadata actionMetadata, TextWriter statusStream
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.InvokeAction(WorkflowObject metabase, TextWriter statusStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.PendingActions.InvokeActions(WorkflowObject metaDb, TextWriter loggingStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionEngine.RunActionQueue()
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.Workflow.RunWorkflow(WorkflowObject workflowObject, HandleInternalException exceptionHandler)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Configuration.BootstrapExtension.ExecuteWorkflowAction.ExecuteAction(String actionId)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Chainer.Infrastructure.Action.Execute(String actionId, TextWriter errorStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionHelper(TextWriter statusStream, ISequencedAction actionToRun)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionWithRetryHelper(WorkflowObject metaDb, ActionKey action, ActionMetadata actionMetadata, TextWriter statusStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.InvokeAction(WorkflowObject metabase, TextWriter statusStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.PendingActions.InvokeActions(WorkflowObject metaDb, TextWriter loggingStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionEngine.RunActionQueue()
(01) 2011-07-12 17:16:01 Slp: Error: Action “Microsoft.SqlServer.Configuration.BootstrapExtension.ExecuteWorkflowAction” threw an exception during execution.
(01) 2011-07-12 17:16:01 Slp: Microsoft.SqlServer.Setup.Chainer.Workflow.ActionExecutionException: Thread was being aborted. —> System.Threading.ThreadAbortException: Thread was being aborted.
(01) 2011-07-12 17:16:01 Slp:    at System.Threading.WaitHandle.WaitOneNative(SafeWaitHandle waitHandle, UInt32 millisecondsTimeout, Boolean hasThreadAffinity, Boolean exitContext)
(01) 2011-07-12 17:16:01 Slp:    at System.Threading.WaitHandle.WaitOne(Int64 timeout, Boolean exitContext)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Configuration.UIExtension.Request.Wait()
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Configuration.UIExtension.UserInterfaceProxy.NavigateToWaypoint(String moniker)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Configuration.UIExtension.WaypointAction.ExecuteAction(String actionId)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Chainer.Infrastructure.Action.Execute(String actionId, TextWriter errorStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionHelper(TextWriter statusStream, ISequencedAction actionToRun)
(01) 2011-07-12 17:16:01 Slp:    — End of inner exception stack trace —
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionHelper(TextWriter statusStream, ISequencedAction actionToRun)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionWithRetryHelper(WorkflowObject metaDb, ActionKey action, ActionMetadata actionMetadata, TextWriter statusStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.InvokeAction(WorkflowObject metabase, TextWriter statusStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.PendingActions.InvokeActions(WorkflowObject metaDb, TextWriter loggingStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionEngine.RunActionQueue()
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.Workflow.RunWorkflow(WorkflowObject workflowObject, HandleInternalException exceptionHandler)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Configuration.BootstrapExtension.ExecuteWorkflowAction.ExecuteAction(String actionId)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Chainer.Infrastructure.Action.Execute(String actionId, TextWriter errorStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionHelper(TextWriter statusStream, ISequencedAction actionToRun)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionWithRetryHelper(WorkflowObject metaDb, ActionKey action, ActionMetadata actionMetadata, TextWriter statusStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.InvokeAction(WorkflowObject metabase, TextWriter statusStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.PendingActions.InvokeActions(WorkflowObject metaDb, TextWriter loggingStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionEngine.RunActionQueue()
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.Workflow.RunWorkflow(WorkflowObject workflowObject, HandleInternalException exceptionHandler)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Configuration.BootstrapExtension.ExecuteWorkflowAction.ExecuteAction(String actionId)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Chainer.Infrastructure.Action.Execute(String actionId, TextWriter errorStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionHelper(TextWriter statusStream, ISequencedAction actionToRun)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.ExecuteActionWithRetryHelper(WorkflowObject metaDb, ActionKey action, ActionMetadata actionMetadata, TextWriter statusStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.ActionInvocation.InvokeAction(WorkflowObject metabase, TextWriter statusStream)
(01) 2011-07-12 17:16:01 Slp:    at Microsoft.SqlServer.Setup.Chainer.Workflow.PendingActions.InvokeActions(WorkflowObject metaDb, TextWriter loggingStream)
(01) 2011-07-12 17:16:05 Slp: Received request to add the following file to Watson reporting: C:UserscluadminAppDataLocalTemptmpBE77.tmp
(01) 2011-07-12 17:16:05 Slp: The following is an exception stack listing the exceptions in outermost to innermost order
(01) 2011-07-12 17:16:05 Slp: Inner exceptions are being indented
(01) 2011-07-12 17:16:05 Slp:
(01) 2011-07-12 17:16:05 Slp: Exception type: System.Management.ManagementException
(01) 2011-07-12 17:16:05 Slp:     Message:
(01) 2011-07-12 17:16:05 Slp:         Invalid query
(01) 2011-07-12 17:16:06 Slp:     Stack:
(01) 2011-07-12 17:16:06 Slp:         at System.Management.ManagementException.ThrowWithExtendedInfo(ManagementStatus errorCode)
(01) 2011-07-12 17:16:06 Slp:         at System.Management.ManagementObjectCollection.ManagementObjectEnumerator.MoveNext()
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.WMIInterop.NodeToActiveGroup.NodeToActiveGroupCollection.NodeToActiveGroupEnumerator.MoveNext()
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.Cluster.WmiClusterGroup.get_OwnerNode()
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.Cluster.ClusterGroup.get_OwnerNodeName()
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.ClusterConfiguration.ClusterGroupPublicConfigObject.get_ClusterGroups()
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.InstallWizard.ClusterGroupController.LoadGroup()
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.InstallWizard.ClusterGroupController.Initialize()
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.InstallWizardFramework.InstallWizardPageHost.SiteObject(Object pageController)
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.InstallWizardFramework.InstallWizardPageHost.CreatePageController(Type type)
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.WizardFramework.PageHost.EnsurePageControllerExists()
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.WizardFramework.PageHost.get_Controller()
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.WizardFramework.PageHost.CreateView()
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.WizardFramework.PageHost.get_Control()
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.WizardFramework.HostControl.OnSelectedPageChanged(Object sender, EventArgs e)
(01) 2011-07-12 17:16:06 Slp:         at System.EventHandler.Invoke(Object sender, EventArgs e)
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.WizardFramework.UIHost.OnSelectedPageChanged()
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.WizardFramework.UIHost.set_SelectedPageIndex(Int32 value)
(01) 2011-07-12 17:16:06 Slp:         at Microsoft.SqlServer.Configuration.WizardFramework.NavigationButtons.nextButton_Click(Object sender, EventArgs e)
(01) 2011-07-12 17:16:06 Slp:         at System.Windows.Forms.Control.OnClick(EventArgs e)
(01) 2011-07-12 17:16:06 Slp:         at System.Windows.Forms.Button.OnMouseUp(MouseEventArgs mevent)
(01) 2011-07-12 17:16:06 Slp:         at System.Windows.Forms.Control.WmMouseUp(Message& m, MouseButtons button, Int32 clicks)
(01) 2011-07-12 17:16:06 Slp:         at System.Windows.Forms.Control.WndProc(Message& m)
(01) 2011-07-12 17:16:06 Slp:         at System.Windows.Forms.ButtonBase.WndProc(Message& m)
(01) 2011-07-12 17:16:06 Slp:         at System.Windows.Forms.Button.WndProc(Message& m)
(01) 2011-07-12 17:16:06 Slp:         at System.Windows.Forms.Control.ControlNativeWindow.WndProc(Message& m)
(01) 2011-07-12 17:16:06 Slp:         at System.Windows.Forms.NativeWindow.Callback(IntPtr hWnd, Int32 msg, IntPtr wparam, IntPtr lparam)
(01) 2011-07-12 17:19:50 Slp: Watson Bucket 1
Original Parameter Values
(01) 2011-07-12 17:19:51 Slp: Parameter 0 : SQL Server “Denali” CTP3@CTP@
(01) 2011-07-12 17:19:51 Slp: Parameter 1 : System.Management.ManagementException.ThrowWithExtendedInfo
(01) 2011-07-12 17:19:51 Slp: Parameter 2 : System.Management.ManagementException.ThrowWithExtendedInfo
(01) 2011-07-12 17:19:51 Slp: Parameter 3 : System.Management.ManagementException@-2146233087
(01) 2011-07-12 17:19:51 Slp: Parameter 4 : System.Management.ManagementException@-2146233087
(01) 2011-07-12 17:19:51 Slp: Parameter 5 : GatherUserSettings
(01) 2011-07-12 17:19:51 Slp:
Final Parameter Values
(01) 2011-07-12 17:19:51 Slp: Parameter 0 : SQL Server “Denali” CTP3@CTP@
(01) 2011-07-12 17:19:51 Slp: Parameter 1 : 0xE8049925
(01) 2011-07-12 17:19:51 Slp: Parameter 2 : 0xE8049925
(01) 2011-07-12 17:19:51 Slp: Parameter 3 : 0x42B4DED7
(01) 2011-07-12 17:19:51 Slp: Parameter 4 : 0x42B4DED7
(01) 2011-07-12 17:19:51 Slp: Parameter 5 : GatherUserSettings
(01) 2011-07-12 17:20:01 Slp: Sco: Attempting to write hklm registry key SOFTWAREMicrosoftMicrosoft SQL Server to file C:Program FilesMicrosoft SQL Server110Setup BootstrapLog20110712_170537Registry_SOFTWARE_Microsoft_Microsoft SQL Server.reg_
(01) 2011-07-12 17:20:01 Slp: Sco: Attempting to write hklm registry key SOFTWAREMicrosoftWindowsCurrentVersionUninstall to file C:Program FilesMicrosoft SQL Server110Setup BootstrapLog20110712_170537Registry_SOFTWARE_Microsoft_Windows_CurrentVersion_Uninstall.reg_
(01) 2011-07-12 17:20:01 Slp: Sco: Attempting to write hklm registry key SOFTWAREMicrosoftMSSQLServer to file C:Program FilesMicrosoft SQL Server110Setup BootstrapLog20110712_170537Registry_SOFTWARE_Microsoft_MSSQLServer.reg_
(01) 2011-07-12 17:20:01 Slp: Sco: Attempting to write hklm registry key SOFTWAREWow6432NodeMicrosoftMicrosoft SQL Server to file C:Program FilesMicrosoft SQL Server110Setup BootstrapLog20110712_170537Registry_SOFTWARE_Wow6432Node_Microsoft_Microsoft SQL Server.reg_
(01) 2011-07-12 17:20:01 Slp: Sco: Attempting to write hklm registry key SOFTWAREWow6432NodeMicrosoftWindowsCurrentVersionUninstall to file C:Program FilesMicrosoft SQL Server110Setup BootstrapLog20110712_170537Registry_SOFTWARE_Wow6432Node_Microsoft_Windows_CurrentVersion_Uninstall.reg_
(01) 2011-07-12 17:20:01 Slp: Sco: Attempting to write hklm registry key SOFTWAREWow6432NodeMicrosoftMSSQLServer to file C:Program FilesMicrosoft SQL Server110Setup BootstrapLog20110712_170537Registry_SOFTWARE_Wow6432Node_Microsoft_MSSQLServer.reg_
(01) 2011-07-12 17:20:06 Slp: Invalid query
(01) 2011-07-12 17:20:06 Slp: Watson bucket for exception based failure has been created
(01) 2011-07-12 17:20:06 Slp: WER: Application level consent value ‘True’ was mapped to consent status ‘WerConsentApproved’
(01) 2011-07-12 17:20:06 Slp: WER: Result of the submission:: ‘WerDisabled’
(01) 2011-07-12 17:20:06 Slp: WER: Submitted 1 of 1 failures to the Watson data repository
(01) 2011-07-12 17:20:06 Slp:
(01) 2011-07-12 17:20:06 Slp: ———————————————————————-
(01) 2011-07-12 17:20:06 Slp:
(01) 2011-07-12 17:20:06 Slp: Error result: -1131919710
(01) 2011-07-12 17:20:06 Slp: Result facility code: 1160
(01) 2011-07-12 17:20:06 Slp: Result error code: 18082

This means that any resource group in the Windows failover cluster that has a backslash (or any other possible illegal character that Setup does not allow) will cause Setup to crash every single time on a first node install. Fun!

There are other potential implications such as the Add Node operation failing (something you definitely don’t want to happen) and possibly uninstall as well (I haven’t tested that one yet). Bottom line: avoid the backslash like the plague. It became almost a running joke for a month or two with the dev guys. A lot of my issues stem from this one cause.

I guess you are wondering why this now does not work if it did in both SQL Server 2008 and SQL Server 2008 R2. The reason is fairly simple: there has been a change in the underlying Setup code that relates to WMI, making a backslash an illegal character during Setup even though it is legal in Windows. I’ve bugged this in a build prior to CTP3 and hope they fix it by the time RTM happens. I’m not sure it’ll be possible. If you feel it’s important, too, let them know!

Popular Posts

Categories