Oracle Observations

November 20, 2011

Wednesday – UKOUG 2011 TEBS planned agenda day 4

Filed under: Uncategorized — bigdaveroberts @ 5:22 pm

Thankfully everyone plans to make it into the conference for 9am on Wednesday!

9:00 – 10:00

Niall plans to attend Redo – presumably a repeat of the session from OAK table Sunday.

Boneist plans to attend Statistics on Partitioned Objects, presented by Doug.

While I approve of both the above choices, I plan to attend Oracle 11g Database: Automatic Parallelism – it sounds like something I need to worry about and it’s by Mr Goodman, Oracle educator par excellonce!

10:10 – 10:55

Niall and Boneist plan to see Using Oracle GoldenGate to Minimize Database Upgrade Risk.

Doug has chosen About multiblock reads, a poor choice if you ask me, as if you turned up on time on Sunday you could already have attended this session!

I’ll probably attend Best Practices for Managing Oracle WebLogic Server.

11:20 – 12:20

Doug and Boneist will attend Latch Internals in 2011.

Niall will attend Oracle Database Smart Flash Cache Overview.

There are several good sessions here! Not highlighted by Doug, Boneist or Niall: Amazon RDS, EC2 and S3 for Oracle Databases and Join Techniques.

However having read the session overviews, I’ll probably do the Latch Internals 2011 with Doug and Boneist.

12:30 – 13:30

A three way split, which is unsurprising given the competition.

Niall votes Oracle RAC One-Node: What, Why, & How.

Doug Votes PGA Memory Management Revisited.

Boneist votes Avoiding Execution Plan instability with SQL Plan Baselines.

Of the above, I’d choose PGA Memory Management Revisited, However I plan to attend the Optimizer Roundtable, but I’d also mention ‘Practical uses for the APEX Dictionary’ that I’ve previously seen at the APEX SIG and is potentially useful if APEX is your thing.

14:25 – 15:10

Niall votes Six Easy Pieces: Essentials of Database Tuning for Packaged Applications.

Boneist votes for this and also votes for Faster, Leaner, More Efficient databases – Index Organised Tables.

Doug neglects to vote.

I’ll be attending the Migrating 1500 Forms modules to Apex session!

15:20 – 16:05

Niall goes with Simulating Workloads with RAT and Simora.

Doug and Boneist vote for Is This The Future of Load & Performance Testing?

My vote goes with Oracle Linux and the Unbreakable Enterprise Kernel: Status Update and Overview.

16:15 – 17:00

For the final session this year Doug selects Transaction Analysis & Diagnosis using Profiling.

Boneist selects Nialls’ session What shape is your data?, which Niall is also attending.

Having attended the last ever UNIX SIG at which Niall presented ‘What shape is your data?’ as the last ever UNIX SIG presentation, I thoroughly recommend this session. OTOH, as I’ve previously supped at this fountain already, I’ll be at the other session with Doug!

Then it’s a visit to my favorite hostelry, that always has Herefordshire Pale Ale on tap and then the trip home.


Tuesday. – UKOUG 2011 TEBS planned agenda day 3

Filed under: Uncategorized — bigdaveroberts @ 12:30 am

9:00 – 10:00

Strangely neither Doug or Boneist appear to be attending the first session of the Day.

Niall indicates his presence at the RAC Attack at certain points during the day, but will he be there at 9:00 am?

I’ll probably attend the Oracle GlassFish Server, Java Enterprise Edition and JVM Roadmap, obviously presuming I’m not with Niall and Doug the previous evening.

10:10 – 10:55

Doug is attending Who’s afraid of Analytic Functions? Which also receives a second vote by Boneist.

Boneist is also attending Nialls’ First Things First: Getting the basics right.

I’m torn between Nialls’ session and the APEX Roundtable. I may choose the roundtable on the basis that with Nialls’ commitment to the Oracle SIGs, I may get the opportunity to revisit his session at another time!

11:15 – 12:15

Niall votes for Troubleshooting RAC.

Doug votes for Execution Plan Interpretation which also receives a second vote from Boneist.

Boneist votes for Analysing Your Data with Analytic Functions.

While I believe I have a good understanding of Execution Plan Interpretation, Past experience of Joze Senegacni indicates that with humour you tend to not get quite what you expected from the session title, and I hope to be presently surprised again!

12:25 – 13:25

Niall and Doug both vote for Instrumentation – Why You Should Care.

Boneist votes for Oracle Optimizer – Best Practices for Managing Optimizer Statistics, Tuning, Refactoring, and Instrumenting Code for Developers and Troubleshooting the Most Complex Oracle Performance Problem I’ve Ever Seen.

My preference is almost always Taniel Poder! and while the session title makes me wonder if I will be able to widely apply what I learn here, I don’t think I’ll regret it!

14:40 – 15:25

Niall, Doug and Boneist have all voted for ASH Outliers: Detecting Unusual Events in Active Session History.

I have never heard of John Beresniewicz. This might be something I have to remedy before the conference!

However unless I’m interested by what I find in the meantime, I’ll be attending the Development tools roundtable!

15:35 – 16:20

Doug and Boneist will be attending Can people be identified in the database?

Niall will be attending Oracle on the Amazon Cloud – What should you consider.

I will be joining Niall in this case.

I should note that Pete Finigans’ session is excellent! I’ve already seen it twice! Pete planned to present this session in 2010 but was prevented by the snow. (Pete isn’t a man of great stature.) He has subsequently presented this at at least 2 SIGs and I strongly recommend this session in spite of the fact that I will not be there!

16:40 – 17:40

Niall votes My Case for Agile Methods.

Doug and Boneist vote for Beating the Oracle Optimizer.

I like to attend Mr Trichlers OTN hands on workshops. I also want to attend Beating the Oracle Optimizer and Moving from Oracle Forms to Oracle ADF: Case Studies.

17:50 – 18:35                                                                                         

Niall, Doug and Boneist are all attending Challenges and Chances of the 11g Query Optimizer.

Simply, Johnathon Lewis recommends Christian Antogninis’ book on the subject of performance troubleshooting.

There can be little higher recommendation!

I will however be attending SQL Developer Data Modeller as a replacement for Oracle Designer.

Hopefully, that will end early and then I’ll sneak into the end of the OTN Workshop!

November 19, 2011

Monday – UKOUG 2011 TEBS planned agenda day 2

Filed under: Uncategorized — bigdaveroberts @ 10:53 am

So Sunday, with at most 2 simultaneous sessions has already left me with one dilemma, obviously Monday brings several more!

10:00 – 10:45

Doug, Niall and Boneist all vote Kyle Hailey.

My vote would be Duncan mills – debugging ADF applications, however as this is an extra length session, it then clashes with other sessions.

11:05 – 12:05

Niall and Boneist vote Connor McDonald – Partitioning 101.

Doug votes Greg Rahn – Deep dive into the SQL Monitoring report.

As stated previously, I’m, inclined to Duncan Mills that  due to its length also clashes with both these sessions.

My natural preference would either be Greg Rhan or Grant Ronalds ‘Future of Forms’.

I suspect that it will be Duncan mills till 11:05, if the session is compelling I’ll stay, if not then depending on how good Connors OAK table session was on Sunday, I might try that or Gregs’ session.

I’m hoping that Grants session overlaps heavily with sessions I’ve done previously. I’ll download the slides though, just in case!

12:15 – 13:15

Doug and Boneist vote Maria Colgan – Upgrading to Oracle Database 11g without pain.

Niall votes Mariusz Piorkowski – Going deeper into Real Application Testing.

While both sessions sound interesting, my organisation is already on 11g, and the performance issues we experience are predominantly application design based issues. I’m not aware of any specific 11g performance issues beyond the fact that the  /*+rule hint tends to be substantially less beneficial in more cases than in the past.

So, I’m probably going for Bryn Llewellyn – Using the PL/SQL Hierarchical Performance Profiler.

14:30 – 15:15

Doug votes Wolfgang Breitling – Using Ashmon.

Niall votes John Hallas – Automatic Diagnostic Repository (Doug also casts a second vote for John).

Boneist votes Simon Haslam – High Availability Options for Modern Oracle Infrastructures. 

Of the above I’d choose Wolfgang, however I’m also tempted by Grant Ronald – ADF’s hidden gem – the Groovy scripting language in Oracle ADF.

Ultimately I’ll probably go to the Application server and fusion middleware Round table. Roundtables are the classic forum for networking with both obscure and interesting real problems!

15:25 – 16:05

Niall votes for MySQL: the least an Oracle DBA needs to know.

Doug has voted for Exadata I/O Resource Management.

Boneist has voted for Implementing Effective Oracle Database Auditing in 3 Steps.

I’ve already had the pleasure of experiencing Patricks’ MySQL session at a SIG. While he promised that he would extend the presentation for the conference, I’m satisfied with what he’s taught me about MySQL already.

And generally any session by an employee of Pythian is worthwhile.

However I plan to attend the ADF Patterns for Forms Conversions by Robert Nocera.

16:35 – 17:35

Both Niall and Boneist vote for Database I/O Performance: Measuring and Planning.

Doug votes for Adaptive Cursor Sharing – An Introduction and again takes a second vote for The DB Time Performance Method (part 1).

While Mr Gorbachev is always insightful, I’m probably going to vote with Dougs’ second choice for the DB Time performance method.

17:45 – 18:30

Niall and Boneist both vote for Doug who will be presenting his Performance and Stability with Oracle 11g SQL Plan Management.

Unfortunately The DB time Performance Method is a double length session and I am unable to attend Dougs’ session which would definitely be my second choice!

November 17, 2011

Four (good) days in December

Filed under: Uncategorized — bigdaveroberts @ 11:39 pm

Obviously I refer to the UKOUG 2011 TEBS Conference!

It is again that time of year, with the agenda published, when we begin to prepare.

Apart from obviously reviewing the now available online conference agenda, several people each year post online their own personal planned agendas.

Currently I have found three agendas posted.

Doug Burns has posted his planned agenda here:

Niall Litchfield has presented his proposed agenda here:

And Boneist has done likewise here:                                                                             

So I have 3 sources to compare my preferences and whims against!

So to my choices, compared with the experts choices:

Day one.

Very pleasingly, there is a welcome return of the Oak Table day!

This occurs on Sunday and has limited places available with a requirement that the day is only available to people who have booked all three days at the main conference.

The first session:

12:15 – 13:15 Mogens Norgaard vs Jonathon Lewis.

Pleasingly both Niall and Boniest agree with my preference for Mr Nogood, Doug unfortunately doesn’t express a preference.

The basis of my decision is based partly on availability/access. While Jonathon makes multiple SIG presentations over the year that overlap with the topics he presents at the UKOUG conference, presentations by Moans are rarer.

This is reinforced by the occasions where Moans attempts to burn down his house, where he still attended the conference but failed to present.

Although he still made his customary appearances at all bar one!

Ultimately, while Jonathans’ presentation will make me a better DBA, Moans presentation will be more entertaining!

13:25 – 14:10 Frits Hoogland vs. Martin Widlake.

Neither Doug nor Boneist have indicated a preference for this slot.

Niall has voted Martin

While I spent an extremely pleasant evening with both Niall and Martin (amongst one other) at Birminghams’  best real ale establishment during last years’ UKOUG conference, I’m voting Frits!

While I appreciate the value of IOTs, I’m currently working on a very well established system and I know that there are substantially less exotic ways to optimise our systems before adopting IOTs.

OTOH, multiblock reads have apparently changed and this is the subject that is most likely to impact me in my current role!

14:35 – 15:20 Christian Antognini vs Connor McDonald

The first session voted on by Doug is for Connor.

Boneist votes for Christian and Niall votes for Connor.

I have attended sessions by Christian before. I have his book. Jonathon Lewis recommends his book. This will undoubtedly be an excellent session.

OTOH, apparently Connor is from Australia and his presence at the UKOUG conference is a rare event.

My vote currently is with Connor, only because of the votes by Doug and Niall although I’m open to persuasion!

15:30 – 16:15 James Morle vs Wolfgang Breitling

Doug, Niall and Boneist all vote for Wolfgang.

That was also my initial inclination.

The first ever presentation that I attended at a UKOUG conference that left me dumbstruck was by Wolfgang. Before that session I had no convincing explanation as to what the cost of a query was in Oracle explain plan terms.

Wolfgang is one of those giants on whose shoulders,  Johnathan Lewis (amongst others) stand upon.

James Morle, OTOH has a breadth of knowledge that can be extremely insightful in the real world.

This is my fist dilemma, as yet undecided!

Blog at