28th February 2007

Flashback database

One of the new features of 10g that we really like is Flashback Database. We had just enabled this feature on a development database and a few days later one of our developers reported that a table had been dropped accidentally. Wow, in the past this would have been a lot of work to recover the dropped table. With flashback database we were able to flashback the tablespace in three minutes (from request to completion). Now that is a really cool time saver.

You will still want to encourage your users to use common sense but flashback will make all our lives easier.

Here is an example implementation step to enable and use flashback.

shutdown immediate;
startup mount;
alter database flashback on;
select flashback_on from v$database — verify it
alter database open;
select flashback_on from v$database — verify it
create restore point before_upgrade;

If you need to restore:

shutdown immediate;
startup mount;
flashback database to restore point before_upgrade;

You can check the restore point with v$restore_point.

We will be writing more about flashback database and other oracle technologies in the future.

posted in podcasting | 0 Comments

8th February 2007

difference between belongs_to and has_one

So this seems to be the difference between belongs_to (in the ActiveRecord DB descriptor files) and has_one
belongs_to means: a field (or fields) in THIS table is a foreign key 
(contains the primary key of) THAT other table.

has_one means: there is a row in THAT table which contains THIS tables 
primary key. [just a row, not necessarily the primary key row over there --that's weird we just use indexes, so don't use.]

Now what I really wish is that ruby would just auto-detect these I don’t know why it wouldn’t. Oh well. It isn’t perfect but it is pretty nice.

posted in podcasting | 0 Comments

7th February 2007

Another test entry

This is another test entry on the way to the blogger for something to see and edit. I will find some informantion on Oracle Open World Oracle search engine expands access to third-party data sources Well that is not really from the openworld but it is a cool post for Secure Enterprise Search.



  1. a numbered list
  2. and another item
  3. last one


This seems to be able to format stuffs his is a testing post for the oracle blog. This blog post is just a test and has nothing here at the moment. So this is the updated part that I have just edited in Elicit and it seems really cool. I would like to see how well the linker and other things work tohis is a testing post for the oracle blog. This blog post is just a test and has nothing here at the moment. So this is the updated part that I have just edited in Elicit and it seems really cool. I would like to see how well the linker and other things work tohis is a testing post for the oracle blog. This blog post is just a test and has nothing here at the moment. So this is the updated part that I have just edited in Elicit and it seems really cool. I would like to see how well the linker and other things work tohis is a testing post for the oracle blog. This blog post is just a test and has nothing here at the moment. So this is the updated part that I have just edited in Elicit and it seems really cool. I would like to see how well the linker and other things work tohis is a testing post for the oracle blog. This blog post is just a test and has nothing here at the moment. So this is the updated part that I have just edited in Elicit and it seems really cool. I would like to see how well the linker and other things work to


Now some other edits are added to this post.


posted in Oracle | 3 Comments

5th February 2007

RadRails in Windows

So Radrails in Windows is fairly nice–you can run your rails commands from a GUI, and it is integrated with the eclipse editor.  Non-trivial to setup in Windows, though–

-e:3:in `load’: C:\InstantRails\ruby\bin\rake.bat:1: parse error,
unexpected tIDENTIFIER, expecting $ (SyntaxError)
from -e:3

means that you haven’t setup the preferences to point to the right rake file, and to do so you may even have to “gem update rake” so that the right rake file (c:\ruby\bin\rake) exists. Yeesh.

posted in podcasting | 1 Comment

5th February 2007

Setting up Ruby on Rails in Windows

Turns out to be somewhat non-trivial :)

http://railsforum.com/rss.php?tid=795

To use the ‘nice’ editor RadRails (replacement for command line Rails–doesn’t do much more than give you boxes instead of making you type it in) and XP you need to follow http://railsforum.com/rss.php?tid=795

Thankfully on Windows there is an installer which gives you most of Ruby for free.

sqlitee seems super trivial to use.

Install it, copy the sqlite3.exe and sqlite3.dll to the ruby\bin folder, and

Just change database.yml to say

development:
adapter: sqlite3
database: cookbook2_development
…repeat X times.

And it should work with normal migrations.

Migrations: so…nothing on the web seems to tell you how to actually start migrations. To do so you run “rails generate migration migrationName” or something along those lines. It will generate it for you, then, with that filename as part of a contention (name db/migrate/migrationName_number.rb) you can put things to setup your DB in that file, like

def self.up
create_table :categories do |t|
t.column :name, :string
end

create_table :recipes do |t|
t.column :title, :string
t.column :value, :text
t.column :description, :string
t.column :date, :date
t.column :instructions, :text
end
end

That type of thing.

Then run rake db:migrate and it learns from the db what ‘step’ of the migration it is at, and updates it automatically to the latest.

So that’s how to setup a migration within Rails. Super easy. It is :)

posted in podcasting | 1 Comment

2nd February 2007

test entry for podcasting

This a test entry for podcasting. This is just a post that has some random information about stuff and not really anything here is actual data. Several people have expressed interest in the automatica archiving of data that i sput into their desk. This will help propel the information age as the fires of doubt will push all data into monolythic repositories first, then like a massive swinging pendulum the data will evacuate to the highly distributed “virtual disk farm” where a file may reside on any or all of thousands of disks. See here


This will cause the data to be highly available and to automatically migrate to the edge of deliver while still maintaining physical, logical and security demands. One file may, for example, demand east and west coast presence for disaster recovery. blah..

posted in podcasting | 0 Comments