October 3, 2011 at 11:23 pm
Comments posted to this topic are about the item How do you learn the advanced stuff?
October 4, 2011 at 12:16 am
I find the biggest constraints to be time and available hardware. One would also need a sample database(s) just to speed things up.
Replication, Failover technologies and disaster recovery all needs sufficient available hardware and this is not always available to the novice DBA that wants to try thing on his own.
October 4, 2011 at 2:17 am
Even working on a 'small' set of data with 1 table of 15 million rows, the fact it was doing lots of user string pattern matching throughout the data and lots of reporting by different users, partitioning brought a 50* performance increase in queries.
So it's definitely worth learning, multi-billion row datasets aren't essential to get benefits from it. Anything where you can greatly reduce the dataset to be queried or modified when using intensive queries can benefit from it.
October 4, 2011 at 3:15 am
Hi Steve,
I don't normally post on the forum (tho read the newsletter daily) and have to say this was a good editorial, it's certainly got me thinking about ways I can get to the advanced topics (especially considering my workplace is considering how to implement advanced reporting and no one has stepped up to plate yet).
October 4, 2011 at 3:47 am
Building and administering Dev Edition on a VM isn't just for advanced features. It's how I learned the simple basics, too.
Very useful technique.
I also keep a database called ProofOfConcept available as a playground for testing new things. Dropable at a moment's notice. Any time I'm testing new concepts, I use that database. I also use it to build answers to questions in these forums. If someone posts table scripts and all that, I fire them up in that database, and build the solution there before posting it here. Very handy, since I know I can drop anything in there at any time, including the whole database.
- Gus "GSquared", RSVP, OODA, MAP, NMVP, FAQ, SAT, SQL, DNA, RNA, UOI, IOU, AM, PM, AD, BC, BCE, USA, UN, CF, ROFL, LOL, ETC
Property of The Thread
"Nobody knows the age of the human race, but everyone agrees it's old enough to know better." - Anon
October 4, 2011 at 3:54 am
I'm busy building a monster of a server (8 cores, 16GB memory, 6TB drive space) that can support multiple virtual machines, so I can play with mirroring, replication, AlwaysOn, clustering and other fun features (there are other reasons for the server too).
Wasn't that expensive. Excluding the drives and screen, under R7000.
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
October 4, 2011 at 4:13 am
If you aer getting a 'developed world' level salary, then probably the best GBP £200 you can spend is on a TechNet subscription. This provides you with all the Microsoft operating systems, server software, and Office. This combination is key to taking hold of your career path. (MSDN can give you more, but at a higher prioce.)
If you think that as a DBA you do not want to know how to install W2008 or configure AD, then you are also missing out on the skills you will need to troubleshoot problems in these areas.
I congratulate Gail on building a Hyper-V monster for trying out what the software can do. This approach liberates you from what an employer can provide, and is another key step in controlling your career.
Back in 2007 I had a 4-core 12 GB box built for about GBP 1800 which now has about 3TB storage spread over 5 disks. If you have got the money, then look at getting a custom-build box with 16 GB memory and at least 8GHz of total processing power (Number of cores x processor speed = total GHz processing power). My experience is that memory and disk space get used up far faster than CPU capacity, and nowadays you should be able to build what you need for under GBP 1000.
Original author: https://github.com/SQL-FineBuild/Common/wiki/ 1-click install and best practice configuration of SQL Server 2019, 2017 2016, 2014, 2012, 2008 R2, 2008 and 2005.
When I give food to the poor they call me a saint. When I ask why they are poor they call me a communist - Archbishop Hélder Câmara
October 4, 2011 at 5:41 am
Thats a reall good article and an issue I habitually have to battle. I go through stages of thinking, I should really learn more about <INSERT FEATURE HERE> but struggle to a) find a block of time to researching it and b) finding a suitable "real world" scenario to apply it to.
I know there are lots of walkthroughs and tutorials out there but I get limited satisfaction from following a set of steps in a pre-canned scenario. Don't get me wrong, I'm not knocking them but I find a more effective learning mechanism in applying the technology to a real problem I have.
Of course, here-in lies the problem and the crux of the editorial. In any single job, you'll be very lucky to have the need to leverage all features of SQL Server even with some creative thinking.
I like the idea of setting aside a couple of hours per week to "test drive" a feature of SQL Server and its something this post re-enforces. Even if you're not getting down and dirty with the feature I think there is value in "kicking the tyres" of something.
Time to block out 2 hours in the calendar...
October 4, 2011 at 6:43 am
Great article. My advice to anyone approaching anything on a computer - whether it's software development, database administration, surfing the web, or just using office productivity applications - is to be curious! In fact it has become my signature tag line.
I also agree with the recommendation for a proper sandbox (invaluable!) and to start with something that's already been done to get you going. I do that all the time. At first, I felt like I was cheating by learning that way. I now look at it as child-like curiosity. You take it apart, see what makes it tick then try to put it back together - and maybe even back together better! It worked for me as a child, and works for my in my current position all the time.
A good manager will serve his/her department well by fostering this within their group. By actually giving their developers/DBA's the time and not just suggesting it, but actually setting it aside for them. Realistically, this is easier for larger organizations with greater depth and overlap of talent.
-- RTW
Be curious!
October 4, 2011 at 7:25 am
I have the same issue that Cheetah does. It comes down to resources, both hardware and time. I love learning but am not able to keep up with day to day responsibities; so I have no choice but to learn only those things that allow me to be productive for my company. Also, we don't have the production hardware we need, much less an adequate test and development environment. I think that may change as we are in the process of virtualizing production servers, which in turn should free up some server we can use to virtualize a test environment.
I do admire anyone's willingness and ability to learn new things ahead of the need, however!
🙂
October 4, 2011 at 9:09 am
In 12 years,(working for a 200 people IT company, and then a 4000 employee non-profit company) there's one thing I've learned. When you die, the last thing you will say is, "I should have spent more time at the office." I used to spend a lot of time reading DBA books and programming. Then I realized I only used about 2-5% of what I actually read. Clustering, partitioning, service broker fall into the area of things I read but probably will never use as a DBA. For me there is no reason to learn it, as I won't use it anytime soon. If I would implement such things, there would be no one but me to support it.
These things will likely change by the next version of Windows/SQL. For me, to read up and be aware of the technology is the only justifiable use of my time. I no longer actually walk through the process of doing unless there is an immediate need. If you need to learn it for an interview, now that's a different story I suppose, but even then with the interviews I toyed with were more interested in real life experience than what was walked through in a book.
If I used the basics to moderate stuff, at least the system admins could help out when I am unavailable if I walk them through it. So, I disagree that these are things every DBA should know. In my opinion, a DBA should know how to do .NET programming instead of learning advanced SQL stuff that you would need a instruction list to follow anyways, as the combination of SQL and .NET skills is very valuable. If you are going to learn advanced stuff, hopefully there is a real purpose for it. Otherwise, you are just making other people's wants your own.
October 4, 2011 at 9:53 am
GilaMonster (10/4/2011)
I'm busy building a monster of a server (8 cores, 16GB memory, 6TB drive space) that can support multiple virtual machines, so I can play with mirroring, replication, AlwaysOn, clustering and other fun features (there are other reasons for the server too).Wasn't that expensive. Excluding the drives and screen, under R7000.
Gail, Would you please send me the specs for this server? I'd like to see what it would cost me to try and put together a server like that one.
October 4, 2011 at 10:16 am
Lynn Pettis (10/4/2011)
GilaMonster (10/4/2011)
I'm busy building a monster of a server (8 cores, 16GB memory, 6TB drive space) that can support multiple virtual machines, so I can play with mirroring, replication, AlwaysOn, clustering and other fun features (there are other reasons for the server too).Wasn't that expensive. Excluding the drives and screen, under R7000.
Gail, Would you please send me the specs for this server? I'd like to see what it would cost me to try and put together a server like that one.
Coolermaster Silencio 550 Sound Proofing Side Panels Piano Black Case
Coolermaster eXtreme Power Plus Dual 12V rail 600w PSU FRPS-C600X
Gigabyte H67MA USB3 B3 LGA 1155 Socket Motherboard REGA-H67MA-USB3-B3
Intel Sandy bridge lga1155 i7-2600(3.4ghz) Quad core Processor
2x Corsair Vengeance Intel XMP 8GB(2x4GB) DDR3-1600 CL9 Desktop Memory Kit FRME-C4G3V16X9x2
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
October 4, 2011 at 11:36 am
Cheetah (10/4/2011)
I find the biggest constraints to be time and available hardware. One would also need a sample database(s) just to speed things up.Replication, Failover technologies and disaster recovery all needs sufficient available hardware and this is not always available to the novice DBA that wants to try thing on his own.
You can do all of this, including shared storage clustering, in VMs. You can run all these features on 1GB VMs (or most of them). They might run slow, but you are testing and learning, and slow is a good thing then.
October 4, 2011 at 11:36 am
Nick Parker-493169 (10/4/2011)
Hi Steve,I don't normally post on the forum (tho read the newsletter daily) and have to say this was a good editorial, it's certainly got me thinking about ways I can get to the advanced topics (especially considering my workplace is considering how to implement advanced reporting and no one has stepped up to plate yet).
Thanks for the post and the kind words. They are appreciated.
Viewing 15 posts - 1 through 15 (of 16 total)
You must be logged in to reply to this topic. Login to reply