Search

Top 60 Oracle Blogs

Recent comments

April 2010

On the Difficulty of Data Migrations (Especially to NoSQL Databases)

(Originally posted at Pythian Blog)

I’ve been reading a lot of NoSQL blogs recently, and one thing that bothers me is that many of the leading NoSQL bloggers seem to have very different experience in operations that I’ve had.

Here’s an example:
Over at the O’Reilly community blogs, Andy Oram interviewed two MongoDB experts about migrating from a relational databases to MongoDB.

Here’s what the experts said:

Slightly OT - How to Pour A Perfect Beer: Duvel

Media: 
See video

DW Introduction

Greg Rahn has been writing a short series on “Core Performance Fundamentals of Oracle Data Warehousing”. Here’s his catalogue of the first four or five articles in the series.

New Oracle OakTable Website

Just a short note to say the Oracle OakTable have launched a wonderful new website: http://www.oaktable.net/. Not only does it look great, but among its many features is a latest news section based on blog feeds from fellow OakTable members, new technical articles, clips and videos, latest twitter updates and is simply a great single location to find [...]

RuleGen

New Features, New Defaults, New Side-Effects

11.2 provides a new feature, deferred segment creation, which is the default for Enterprise Edition databases, whilst this feature is rather nice in some restricted circumstances, there are a couple of side effects which may catch the unwary and in some cases make previously valid operations significantly more difficult.

Did you know...

It is Saturday (ok, you probably knew that) so I'll do something non-Oracle. Did you know - carrots aren't really supposed to be orange? They should be purple. And they are if you grow a few generations of the Orange ones (they revert back to their natural state).

So, much like the yellow, easy to peel bananas we enjoy - Orange carrots are a product of "us".

I wonder what Bugs Bunny would have to say about that...

My database is slow - now what ?

Although this example was supplied as a simple template, it's worth using it as a link point to an OTN post where Randolf Geist explains how to ask for help when trying to sort out performance problems with a single SQL statement.

Latch - cache buffer chains on small index (primary key) caused by concurrent batch scripts (select statement) around 300

The Oak Table challenge is not currently live - it goes active when a group of members of the OakTable arrange to meet at a public event - but since that wasn't clear when you raised the question, here are a couple of thoughts.
 
First - The Oracle version number really matters in this one; there have been bugs with latching on the root block of an index - the most recent that I know of being when an index had been rebuilt. See this blog item  
 
Second - is the latching problem on the root block or on the leaf blocks ? If it's only on the root block then trying to spread the index entries across multiple leaf blocks isn't going to help anyway
 
Third - Using pctfree to limit the number of rows per block only works when you create or rebuild the index; if you're deleting and inserting rows in the batch then the pattern of activity may simply leave you with well-packed blocks after a very short time. (Possibly this is why you are rebuilding the index every two minutes - but that's likely to cause other, more significant, overheads anyway).
 
Fourth - your question suggests that  you're only using select statements against this table: if that's true I can't see how rebuilding the index every two minutes could help - unless you're on an old version of Oracle that had bugs that generated excessive read-consistent copies of blocks.
 

Oaktable Website

Thanks to the valient work of James Morle there is a awesome new website aggregating the work and postings of the members of the Oaktable network. Check it out: