Month: March 2015

PDB Logging Clause : What you give with one hand, you take away with the other…

The Oracle 12c 12.1.0.2 release came with a bit of functionality called the PDB Logging Clause. Or to put it another way, it didn’t because the feature just didn’t work. I raised an SR highlighting the issue.

The PDB logging clause is meant to allow you to define a default logging clause for the whole PDB. If a tablespace is created in the PDB without an explicit logging clause, the current PDB logging clause setting should be used.

It was listed as a new feature in 12.1.0.2, but like I said, it just didn’t work. You could alter the PDB logging clause setting, but the CREATE TABLESPACE command seemed oblivious to this setting…

I wrote an article about the feature, with a big note about it not working, because it confused the hell out of me when I was looking at it and I saw a bunch of other blogs talking about the feature, without mentioning it was actually broken. I figured it might be worth setting the record straight for any confused parties out there.

Today (6 months later) I got an update to the SR telling me the issue had been fixed by patch 18902135. I downloaded the patch and applied it to a test instance. At first it looked like it was working, but it has introduced a new problem. Once patched the functionality does the following.

  • If the PDB logging clause is set to NOLOGGING, any new tablespaces created without an explicit logging clause are created as NOLOGGING. That sounds like a step in the right direction! It didn’t do this before the patch was applied. 🙂
  • Unfortunately, you *can’t* override this by explicitly specifying the logging clause in the CREATE TABLESPACE command.

The docs are quite clear about how it should work.

“The specified attribute is used to establish the logging attribute of tablespaces created within the PDB if the logging_clause is not specified in the CREATE TABLESPACE statement.”

Fortunately, there is a workaround, which is to assume the feature doesn’t exist and keep doing things the way you always have. That is, manually define tablespaces as NOLOGGING when you need them.

I’ve added the problems associated with the current “bug fix” to my SR and this has been referred back to development. Let’s see how long the fix for this takes… 🙂

Cheers

Tim…

Oracle Midlands Event #8 : Summary

Last night was Oracle Midlands Event #8 sponsored by Redgate.

There was a problem with the projector, but fortunately there were a couple of large TV screens, so it didn’t have to be a complete no-slide zone. Neither of the talks relied on displaying lots of code, so I don’t think this caused a big problem.

First up was Chris Saxon speaking about Edition-Based Redefinition. I’ve done talks on this subject, so I was interested to see how Chris approached it. The talk started with an explanation of the problems associated with deploying new versions of PL/SQL code to production, then moved on to possible solutions available prior to 11gR2. Once that ground work had been established, he moved on to explain how EBR can be used to make the process more robust, focussing on some specific pieces of the EBR functionality. I preferred his approach to the subject than my own, so I was taking a bunch of notes about his presentation style. There is always something to learn. I thought it was a really good session and Chris handled the subject (and the projector issue) really well. His recent move to Steven Feuerstein’s advocacy/evangelist team at Oracle means I should get to see him speaking some more in the future, which will be cool.

After the food break, it was my session on “Pluggable Databases – What they will break and why you should use them anyway!” It was the first time I’ve done this session in front of a crowd, which is always a bit nervy. It seemed to go down pretty well. Here is the feedback from those who filled in the evaluations.

  • “Good presentation”
  • “Good presentation, dynamic material & delivery”
  • “A lot of content, probably requires a part 2”
  • “Scary!!!”
  • “Good information and entertaining delivery style”
  • “Excellent – superb topic & presentation skills”
  • “Very good talk. Much information gained”
  • “Brilliant & scary!”
  • “Informative, passionate & useful”
  • “Brilliant”
  • “Really good insight/information. Real world understanding which makes a real difference. As always brilliant!”
  • “Not bad for a beginner”
  • “Really enjoyed & challenging subject matter”
  • “Fantastic!”
  • “Good”

The, “Not bad for a beginner”, comment made me LOL when I read it. There are a couple of people I’m putting in the frame for that one. 🙂

It wasn’t my intention to make the Mulitenant option seem really scary. I thought I was doing a sales pitch for it, not scaring people off. Maybe I need to re-frame things a little… 🙂

Thanks to Mike for getting the event sorted and thanks to Redgate for the sponsorship. Thanks to Chris for coming along. Hopefully we can get him back again in the future. Big thanks to everyone who turned up to the event last night. Let’s keep this train rolling!

Cheers

Tim…

VirtualBox 4.3.24

VirtualBox 4.3.24 has been released. Downloads and changelog are in the usual places.

Regarding the upgrades:

  • Fedora 21 : No problems.
  • Mac OS X Yosemite : No problems.
  • Windows 7 : I had to use the task manager to kill the “VBoxNetDHCP.exe” and “VBoxSVC.exe” processes from the current installation before the upgrade could continue. Once those were killed, it installed and worked fine.

Happy upgrading!

Cheers

Tim…

The Definition of Success

After reading yet another article alluding to the death of Google+, I’m left pondering the subjective definition of success.

Success : the accomplishment of one’s goals.

Based on that definition, then Google totally failed. I’m sure their aim was to supplant Facebook as the “time-wasting app of choice”, which they’ve certainly not done.

Success : a person or thing that is successful

There are active communities on Google+, with a significantly smaller signal-to-noise ratio compared to the Facebook equivalent. I’m sure the members of those communities will consider Google+ a success.

Rating your own success is also subjective. After nearly three years in my currently role, some would describe my achievements as considerable. By my own estimates, it feels like I’ve achieved very little. It is nice to be praised for work you are proud of, but very hard to accept praise for work that falls beneath the standards or productivity level you expect from yourself.

I’m lucky enough to go round the world giving people advice about how they should be using Oracle products, but the downside of this is you sometimes look at what you are doing at work and ask yourself, “What the hell would people think if they saw this shit?”

I always try to work on the basis I want things to be as simple and clear as possible, so that if I left the company the person that followed me would sit back and think, “Wow. The boy done good!” Unfortunately, a liberal dose of reality and questionable management decisions means you are often left polishing a turd!

As you’ve probably guessed, I’m on a bit of a downer at the moment. I know I’m pretty good at what I do and I get a kick out of the level of productivity I’m capable of, when I’m not being held back. The counter to that is I hate feeling ordinary, which is exactly how I feel today as I plod along at a snails pace, polishing that turd for all it’s worth. It’s totally an ego thing and I need to get over myself, but there it is…

Cheers

Tim… (A decidedly ordinary person doing decidedly ordinary things in Birmingham)