I’ve said a number of times, the process of writing articles is part of an ongoing learning experience for me. A few days ago my personal tech editor (Jonathan Lewis) asked about a statement I made in the SQL Plan Directive article. On further investigation it turned out the sentence was a complete work of fiction on my part, based on my misunderstanding of something I read in the manual, as well as the assumption that everything that happens must be as a result of a new feature.
Anyway, the offending statement has been altered, but the conversation this generated resulted in new article about Automatic Column Group Detection.
The process also highlighted how difficult, at least for me, it is to know what is going on in the optimizer now. It wasn’t always straight forward before, but now with the assorted new optimizations, some beating others to the punch, it is even more difficult. There are a number of timing issues involved also. If a statement runs twice in quick succession, you might get a different sequence of events compared to having a longer gap between the first and second run of the statement. It’s maddening at times. I’m hoping Jonathan will put pen to paper about this, because I think he will do a better job of explaining the issues around the inter-dependencies better than I can.
Anyway, I will be doing another pass through this stuff over the coming days/weeks/months/years to make sure it is consistent with “my current understanding”.
Fun, fun, fun…
As always, installations of Oracle server products on Fedora are not a great idea, as explained here.
I was reading some stuff about the Fedora 23 Alpha and realised Fedora 22 had passed me by. Not sure how I missed that.
Anyway, I did a run through of the usual play stuff.
While I was at it, I thought I would get the heads-up on Fedora 23 Alpha.
The F23 stuff will have to be revised once the final version is out, but I’m less likely to forget now.
I guess the only change in F22 upward that really affects me is the deprecation of YUM in F22 in favour of the DNF fork. For the most part, you just switch the command.
yum install my-package -y
yum groupinstall my-package-group -y
yum update -y
dnf install my-package -y
dnf groupinstall my-package-group -y
dnf group install my-package-group -y
dnf update -y
This did cause one really annoying problem in F23 though. The “MATE Desktop” had a single documentation package that was causing a problem. Usually I would use the following.
yum groupinstall "MATE Desktop" -y --skip-broken
Unfortunately, DNF doesn’t support “–skip-broken”, so I was left to either manually install the pieces, or give up. I chose the latter and use LXDE instead. F23 is an Alpha, so you expect issues, but DNF has been in since F22 and still no “–skip-broken”, which I find myself using a lot. Pity.
My recent dalliance with YouTube (141 subscribers and growing! ) has left the blog feeling a little bit unloved of late, but then why write when you can waffle in the car?
Anyway, the 12c learning train keeps on rolling. I’ve recently put the following articles live.
These are all listed as 12c new features in the 1Z0-060 “Upgrade to Oracle Database 12c” OCP upgrade exam, which I find a bit odd. Two of them are EM12c features, not database features. The other two are existing EM12c features that are now available directly from the database, but I can’t see myself ever using them on the server when it is much easier to do it from Cloud Control. Whatever!
Getting close to the end of the OCP syllabus now… I’ll soon have to consider sitting the exam…
I was looking for something in the New Features Manual and I had a total WTF moment when I saw this stuff.
If you look at the final section of the article, you can see in some cases these just get transformed to regular joins and outer joins, but there is certainly something else under the hood, as shown by the pipelined table function example.
I think it’s going to take me a long time before I think of using these in my regular SQL…
Update: The optimizer has used
LATERAL inline views during some query transformations for some time, but they were not documented and therefore not supported for us to use directly until now. Thanks to Dominic Brooks and Sayan Malakshinov for the clarification.
A little over a year ago I was at the BGOUG Spring Conference and I watched a session by Maja Veselica about auditing in Oracle Database 12c. At the time I noted that I really needed to take a look at this new functionality, as is was quite different to what had come before. Fast forward a year and I’ve finally got around to doing just that.
I’ve tried to keep the article quite light and fluffy. The Oracle documentation on this subject is really pretty good, so you should definitely invest some time reading it, but if you need a quick overview to get you started, my article might help.
My 12c learning experience continues…
I was planning to cover this subject in a single article, but it got a bit bulky, so I split it down into 6 little articles.
I’ve also created a links page to bring them all together.
I guess you could call it a list of nice-to-haves, rather than something revolutionary, but I’m sure someone will come back to me saying one of them has changed their life!
I decided I wanted to play with the newly released Spartan browser on Windows 10. Spartan comes with Windows 10 (build 10049), which does not have an ISO download available at the moment. So instead, I downloaded the x64 ISO image of Windows 10 (build 10041) and installed it on VirtualBox.
To get build 10049 you have to switch the Windows Update settings from “Slow” to “Fast”, which gives you access to the latests builds as soon as they are available.
That done, Windows Update will then download build 10049, which is pretty much a full OS download again. Once rebooted, the OS auto-installs for ages, with a few reboots, but when it is done you are left with the latest Windows 10 build.
It boots to the desktop and feels quite similar to Windows 8.1. If you are interested in the latest start menu, here it is.
If I’m honest, I’ve never seen the Windows 8.1 start menu live. The Windows 8 menu was so bad I installed Classic Shell on the Windows 8 machines for my family. I’ve never removed it since the 8.1 upgrade. As a result, I don’t really know if this Windows 10 start menu is new or like the 8.1 menu. I would probably stick with this menu myself, knowing that Classic Shell is always available if it pisses me off.
Most importantly, THIS IS SPARTA(N)!
Not surprisingly, it’s a just a browser and any site that sticks reasonably close to the standards will work fine.
So that was the fun bit. Now I’ve got to look at what this is going to break. I’m guessing Oracle Forms isn’t going to like it.
PS. Alex – By “and junk” I was not implying it is junk. Este Uimitor!
Update: Installed Oracle Database 12c on Windows 10 without any problems. Happy days!
I keep thinking I’m moving forward with this Oracle database 12c stuff, but around every corner there is another surprise. A few days ago I was setting up a demo for Transparent Data Encryption (TDE) in 12c using my existing articles (10g, 11g). That’s when I noticed things had changed, so I had to use an 11g instance for the demo and make a note to revisit TDE for 12c…
On revisiting the subject, I saw that the encryption key management has changed in 12c. What’s more, if you are using the multitenant option it is a bit different again. That resulted in this article.
While I was working through this I was getting some freaky results, which were driving me mad. Whilst trying to figure out that I noticed I had two PDBs of the same name under a single listener. I had created two test instances (cdb1 and cdb2), each with a PDB called “pdb1”. There is a sentence in the docs to say this is not a good idea, which resulted in this little article.
So it turns out that TDE works fine, provided you are not an idiot.
The moral of the story is RTFM carefully, because sometimes a single sentence can make all the difference!
I’ve been having a play around with the enhancements to the statistics collection in 12c. I’ve put together this top-level post with links to all my other articles on this subject.
Here are the new articles it links to.
It also links to some of the stuff I put out previously for the Adaptive Query Optimization functionality, as that is statistics related.
In a totally unrelated incident, I wrote this thing about the new READ object privilege, but forgot to mention it on the blog.
The journey continues…
Having played around with Flashback Data Archive in 11g, I figured I would get through the 12c enhancements pretty quickly. I didn’t account for the fact I’m a donkey and can’t see the wood for the trees. Luckily, I know some people who aren’t stupid and they gave me a nudge in the right direction, allowing me to spot my silly mistake. Thanks Bjoern and Connor!
So after lots of wasted time, here is the article.
For the most part, it’s an evolution, but the new bits are pretty darn cool. I guess a lot of people will focus on two main things:
- It is now a free feature, provided you don’t use compression, available in all versions. The change to use no compression by default has been back-ported to 126.96.36.199, so it’s free there too, which is nice!
- The contents of contexts can now be stored in the flashback archive, so you can have access to the USERENV and custom context values that were set in the session when the DML was processed. This makes it possible to replace all those crappy old audit triggers with FDA!
There’s some other stuff in there that’s kind-of nice too. I think it’s worth checking it out, especially at its new price.