Still selling well: OCA Oracle Database SQL Exam Guide (Exam 1Z0-071)

My August 2017 book for Oracle Press – an imprint of McGraw-Hill Education – is still selling well.  Amazon this morning ranks it as follows:

Not bad for a 2017 book!  It’s available here:


Python and SQL

Is it me?  Or is there a lot of buzz about Python lately?  Python is a great language for doing certain work quickly.  It supports a variety of software design approaches.  It is interpreted, and yet is object-oriented.  You can use it like an interactive calculator, and it comes with powerful math libraries.  It’s succinct – you can do a lot with fewer lines of code than is typical, but it can be used for large scale programming as well.  For these reasons and more, Python is growing in popularity, particularly in the fields of data analytics and data science.

It makes sense that it works well with SQL to provide data persistence. This blog post shows you how you can leverage existing SQL data stores with the power of Python.

It makes sense that it works well with SQL to provide data persistence. This blog post shows you how you can leverage existing SQL data stores with the power of Python.

Getting Ready

I downloaded and used the following software:

Python 3.5.2 Popular software language!
SQLite 3.15.2 Embedded server-less SQL database
Oracle 12c The world’s leading relational database management system
cx_Oracle  5.2.1 Python extension module

I downloaded and installed Python and Oracle.  I also downloaded SQLite and cx_Oracle for use with Python.

Python and SQLite

SQLite is an embedded SQL tool.  It doesn’t have a separate server process, but is easily incorporated into applications. Here’s an example of Python code that uses SQLite, version 3:

import sqlite3
connection = sqlite3.connect("ships.db")
cursor = connection.cursor();
sql_statement = """CREATE TABLE ships
                  (    ship_id INTEGER
                     , ship_name TEXT);"""
sql_insert = """INSERT INTO ships VALUES(1, 'Codd One');"""
sql_insert = """INSERT INTO ships VALUES(2, 'Codd Two');"""
sql_insert = """INSERT INTO ships VALUES(3, 'Codd Three');"""

The code above will successfully create a SQLite database file “ships.db”, then create table “ships” and then add three rows to it, and COMMIT the records.

In a subsequent script, you can query the same database.

import sqlite3
connection = sqlite3.connect("ships.db")
cursor = connection.cursor();
sql_statement = """SELECT ship_id, ship_name FROM ships;"""
result = cursor.fetchall()
for r in result:

The rows returned by the SELECT statement are processed within the “for” loop.  “Result” captures the result of the query, and “r” serves as a fetched value for each row.  In this case, each row represents an array of two values, and both values are displayed with the “print(r)” statement.

The output:

(1, 'Codd One')
(2, 'Codd Two')
(3, 'Codd Three')

Note:  the “>>>” symbol is the Python cursor, indicating the interactive system is awaiting the next instruction.

Python and Oracle

Now let’s use Python to send SQL statements to Oracle 12c. To establish the connection, we use the Python extension module “cx_Module”, available from – see above for the links to all the software necessary to run these scripts.

Also – note that while SQL strings submitted directly to Oracle require semi-colons, SQL statements sent via cx_Oracle reject them.

In this example, we’ll do the same thing we did with SQLite, with one change: instead of breaking up the CREATE/INSERT and SELECT statements into two scripts, we’ll do it all in one script:

import cx_Oracle
connection = cx_Oracle.connect('username/password@localhost/orcl')
print("Oracle RDBMS version " + connection.version)

cursor = connection.cursor()

sql_statement = """DROP TABLE ships2"""

sql_statement = """CREATE TABLE ships2 
                   (   ship_id NUMBER
                     , ship_name VARCHAR2(20))"""

sql_insert = """INSERT INTO ships2 VALUES(1, 'Codd Crystal')"""
sql_insert = """INSERT INTO ships2 VALUES(2, 'Codd Elegance')"""
sql_insert = """INSERT INTO ships2 VALUES(3, 'Codd Champion')"""


cursor.execute("SELECT ship_id, ship_name FROM ships2")
result = cursor.fetchall()
for r in result:


print("End of script, database connection closed.");

The output of this script is the same we saw earlier:

(1, 'Codd One')
(2, 'Codd Two')
(3, 'Codd Three')

Regardless of which database we use, the results are the same.

Closing Thoughts

In this blog post, we looked at how to use Python to execute SQL statements.  We created SQL objects, and sent data to the database, as well as queried data from the database.  We passed data to and from Python structures and set t he stage for any data manipulation we wish to do using these powerful tools.

Python is easy to use, and capable of performing complex math functions and data manipulation.  Using SQL to persist data for Python is an ideal combination for data analytics.

NOTE: This data and the accompanying scripts are available at my GitHub site:

Number one again!

I was pleased to see that my SQL Expert book is yet again listed as a number one best seller this afternoon by’s book ranking system for two of the book’s categories:

  • Computers and Technology -> Databases and Big Data -> Oracle
  • Computers and Technology -> Certification -> Oracle

Trailing are excellent books in their own right, including many authored or co-authored by Oracle ACE Directors: a brand new Oracle Public Cloud book co-authored by Charles Kim, Mark Rittman‘s Business Intelligence book from Oracle Press, and Steve Feuerstein‘s legendary PL/SQL book.

I’m busy working on a new book from Oracle Press, more on that in the very near future – stay tuned! In the meantime, the 047 book makes a great holiday present for anyone in your life who wants to become a certified SQL Expert!

Below is the screenshot from the “Databases and Big Data” category:

Oracle SQL Expert Number One on the Amazon Best Selling Charts 11/29/2016 for its category.

Rio Mid-Games: Gold Medals Per Capita So Far

Four years ago I blogged about the fact that medal counts are often reported on a per-country basis without regard for the population of any country. I’ve always thought that approach to be silly. The per-capita medal count is much more interesting to me.

So this morning I used Bing to get the latest count by country of gold medals at the ongoing Rio Olympics.  I included only those nations that have won two or more gold medals.

I then visited Wikipedia to get recent population counts for those nations.

I used all this data to determine the per-capita win per nation and ranked them.  The results are below.

Ranking Country Per Capita Ratio
(Descending Order)
Gold Medal Count Population Date Population Reported
1 Jamaica 7.34418E-07 2 2,723,246 31-Dec-14
2 New Zealand 6.37423E-07 3 4,706,450 16-Aug-16
3 Hungary 6.10811E-07 6 9,823,000 1-Jan-16
4 Croatia 4.77251E-07 2 4,190,669 31-Dec-15
5 Netherlands 3.52389E-07 6 17,026,640 16-Aug-16
6 Australia 2.48343E-07 6 24,160,100 16-Aug-16
7 United Kingdom 2.45738E-07 16 65,110,000 30-Jun-15
8 Switzerland 2.39762E-07 2 8,341,600 31-Mar-16
9 Greece 1.84196E-07 2 10,858,018 1-Jan-15
10 Cuba 1.77952E-07 2 11,239,004 31-Dec-15
11 Belgium 1.76637E-07 2 11,322,674 1-Jun-16
12 Italy 1.31871E-07 8 60,665,551 1-Jan-16
13 South Korea 1.18107E-07 6 50,801,405 1-Jul-16
14 Kazakhstan 1.12656E-07 2 17,753,200 1-May-16
15 Germany 1.10064E-07 9 81,770,900 30-Sep-15
16 France 1.04891E-07 7 66,736,000 1-Jul-16
17 Spain 8.61356E-08 4 46,438,422 1-Jan-16
18 United States 8.01855E-08 26 324,248,000 16-Aug-16
19 North Korea 7.91108E-08 2 25,281,000 1-Jul-16
20 Russia 7.50345E-08 11 146,599,183 1-May-16
21 Uzbekistan 6.33406E-08 2 31,575,300 1-Jan-16
22 Canada 5.53166E-08 2 36,155,487 1-Apr-16
23 Japan 5.51225E-08 7 126,990,000 1-Jul-16
24 Poland 5.20329E-08 2 38,437,239 31-Dec-15
25 Kenya 4.52934E-08 2 44,156,577 1-Jul-15
26 Colombia 4.09887E-08 2 48,793,900 16-Aug-16
27 Thailand 3.04279E-08 2 65,729,098 31-Dec-15
28 Iran 2.51689E-08 2 79,463,100 16-Aug-16
29 China 1.08836E-08 15 1,378,220,000 16-Aug-16
30 Brazil 9.68457E-09 2 206,514,000 16-Aug-16

You could argue that some country populations and/or medal counts are too small to prove anything other than an interesting anomaly.  Or not.  But there’s no question that Hungary and the Netherlands are very impressive, each with six gold medals for an overall per-capita rate that’s dramatically higher than, for example, Great Britain, the United States, and especially China, who – in spite of their very high medal count, is actually very far behind the other nations on a per capita basis.

These are the results presented graphically, which I created using Microsoft Excel, captured with the built-in “snipping” tool in Microsoft Windows: :

The Pareto Line seems to curve most significantly above Hungary and the Netherlands, which have the most striking performance in terms of size of population and count of gold medals.  Also – note how far back China ranks.  They might have a lot of medals, but they also have the largest population in the world.

To me, this sort of analysis is much more interesting than the simple medal count that doesn’t account for population.


  • Gold Medal Count as of 8/16/2016:
  • Country Population Data:

MySQL Migration: from Drupal to WordPress

I just migrated this Skere9 blog from a Drupal implementation at SiteGround to a WordPress instance at Amazon Web Services.  I’ve been a customer of AWS for three years and finally decided to move this blog to the same platform where I’ve been doing other work.  It’s cost-effective and AWS provides significantly greater control over numerous aspects of the deployment.

Note:  I have one author – myself – and I did not migrate any comments with this move.  I will add those later.

The Drupal version is 7.4.1.  The WordPress version is 4.5.3.

To migrate the blog posts, I took the following steps.

  • I temporarily installed phpMyAdmin at the Drupal site to export the “node” and “field_data_body” tables from Drupal.  (Note: this could have also been done using MySQL directly.)
  • I installed phpMyAdmin at the WordPress site to import the Drupal tables.  (Note: this could have also been done using MySQL directly.) I edited the phpMyAdmin config file to restrict access to phpMyAdmin to my own IP address.
  • I executed the script below via phpMyAdmin at the WordPress site to pull data from the imported Drupal tables and insert that data into the WordPress database.
INSERT INTO wp_posts
(       id
      , post_author
      , post_date
      , post_date_gmt
      , post_content
      , post_title
      , post_status
      , comment_status
      , ping_status
      , post_name
      , post_modified
      , post_modified_gmt
      , post_parent
      , guid
      , menu_order
      , post_type
      , 1
      , FROM_UNIXTIME(a.created)
      , FROM_UNIXTIME(a.created)
      , b.body_value
      , a.title
      , 'publish'
      , 'open'
      , 'open'
      , CONCAT('dr-', a.nid+100)
      , FROM_UNIXTIME(a.changed)
      , FROM_UNIXTIME(a.changed)
      , 0
      , CONCAT('', a.nid+100)
      , 0
      , 'post'
FROM     node a LEFT JOIN (field_data_body b)
            ON (a.nid = b.entity_id)
WHERE    type = 'blog'
ORDER BY a.nid

A few important notes about the above code:

  • I had fewer than 100 existing posts in the WordPress instance- it was a brand new installation.  This is why I only added 100 to the ID, to create unique identifiers.
  • I transformed the data formats from UNIX to plain text, as required by the source and target tables.
  • I chose to enter each value into the WordPress “postname” field with a “dr-” prefix – for “Drupal”.  But of course that could have been anything, or nothing at all – no prefix is required here.
  • The “_gmt” fields in the target database are being treated by my WordPress implementation as the same as the local time.  Presumably GMT refers to Greenwich Mean Time, comparable to UTC.  I’m in the USA eastern time zone (New York City zone – I’m in Washington, DC), so presumably the offset should be 5 hours but the existing WordPress implementation I have isn’t treating the times differently.  Therefore I didn’t treat them differently either – I made the “created” timestamp the same for the local and GMT entry, and did the same for the “changed” timestamp.

It all worked like a charm.


Here’s something I’ve been working on lately – a number of metrics across large data sets where I want to determine aggregates of aggregates.  I can’t present the actual data I’m working with here, but this is an illustrative example.  Consider the data set below.

  2  FROM     shop_orders;

---------- ---------- --------------------
         1 Office     Stapler
         2 Fruit      Banana
         3 Bread      French Bread
         4 Steak      Ribeye
         5 Fish       Bass
         6 Crackers   Saltines
         7 Fish       Swordfish
         8 Fruit      Pear
         9 Bread      Rolls
        10 Crackers   Chips
        11 Steak      Prime Rib
        12 Bakery     Birthday Cake
        13 Fish       Salmon
        14 Fish       Flounder
        15 Cleaners   Detergent
        16 Bakery     Donuts
        17 Office     Ink Cartridges
        18 Steak      Filet Mignon
        19 Vegetable  Broccoli
        20 Steak      Flank
        21 Fruit      Raspberry
        22 Sundries   Shampoo
        23 Steak      Salisbury
        24 Vegetable  Radish
        25 Vegetable  Mushroom
        26 Pharmacy   Band-aids
        27 Pharmacy   Aspirin
        28 Sundries   Toothbrush
        29 Sundries   Shampoo
        30 Cereal     Bran
        31 Cereal     Corn Flakes
        32 Cleaners   Soap
        33 Cereal     Oatmeal

33 rows selected.

One glance at the data and you can see the category values repeat. But how often do they repeat? Which categories recur the most frequently? What if you needed to determine which categories have the greatest number of values, and which don’t?

To determine this, we can start with a GROUP BY, like this.

SQL> SELECT   category, COUNT(*) CT
  2  FROM     shop_orders
  3  GROUP BY category;

---------- ----------
Cereal              3
Fruit               3
Crackers            2
Vegetable           3
Office              2
Steak               5
Pharmacy            2
Fish                4
Bread               2
Cleaners            2
Sundries            3
Bakery              2

12 rows selected.

Now let’s move the GROUP BY to an inline view, and do a GROUP on the GROUP BY, like this.

SQL> SELECT COUNT(b.category) ct_category, b.ct ct_item_n
  2  FROM
  3    ( SELECT   a.category, COUNT(a.item_n) CT
  4  	 FROM	  shop_orders a
  5  	 GROUP BY a.category
  6    ) b
  7  GROUP BY b.ct
  8  ORDER BY b.ct DESC;

----------- ----------
          1          5
          1          4
          4          3
          6          2

Voila – there is one category that recurs five times, one recurs 4 times, four recur three times, and finally six categories have two values each.

In a relatively small data set like this, the answers might be obvious. I’ve been working with very large rows that number in the tens of millions and queries like this have been invaluable in confirming data shape and patterns.

If you wish to experiment with these SQL statements, here is the code to create the sample table and sample data.

DROP   TABLE shop_orders;
CREATE TABLE shop_orders
(   id       NUMBER
  , category VARCHAR2(15)
  , item_n   VARCHAR2(20)
INSERT INTO shop_orders VALUES ( 1, 'Office', 'Stapler');
INSERT INTO shop_orders VALUES ( 2, 'Fruit', 'Banana');
INSERT INTO shop_orders VALUES ( 3, 'Bread', 'French Bread');
INSERT INTO shop_orders VALUES ( 4, 'Steak', 'Ribeye');
INSERT INTO shop_orders VALUES ( 5, 'Fish',  'Bass');
INSERT INTO shop_orders VALUES ( 6, 'Crackers', 'Saltines');
INSERT INTO shop_orders VALUES ( 7, 'Fish', 'Swordfish');
INSERT INTO shop_orders VALUES ( 8, 'Fruit', 'Pear');
INSERT INTO shop_orders VALUES ( 9, 'Bread', 'Rolls');
INSERT INTO shop_orders VALUES (10, 'Crackers', 'Chips');
INSERT INTO shop_orders VALUES (11, 'Steak', 'Prime Rib');
INSERT INTO shop_orders VALUES (12, 'Bakery', 'Birthday Cake');
INSERT INTO shop_orders VALUES (13, 'Fish', 'Salmon');
INSERT INTO shop_orders VALUES (14, 'Fish', 'Flounder');
INSERT INTO shop_orders VALUES (15, 'Cleaners', 'Detergent');
INSERT INTO shop_orders VALUES (16, 'Bakery', 'Donuts');
INSERT INTO shop_orders VALUES (17, 'Office', 'Ink Cartridges');
INSERT INTO shop_orders VALUES (18, 'Steak', 'Filet Mignon');
INSERT INTO shop_orders VALUES (19, 'Vegetable', 'Broccoli');
INSERT INTO shop_orders VALUES (20, 'Steak', 'Flank');
INSERT INTO shop_orders VALUES (21, 'Fruit', 'Raspberry');
INSERT INTO shop_orders VALUES (22, 'Sundries', 'Shampoo');
INSERT INTO shop_orders VALUES (23, 'Steak', 'Salisbury');
INSERT INTO shop_orders VALUES (24, 'Vegetable', 'Radish');
INSERT INTO shop_orders VALUES (25, 'Vegetable', 'Mushroom');
INSERT INTO shop_orders VALUES (26, 'Pharmacy', 'Band-aids');
INSERT INTO shop_orders VALUES (27, 'Pharmacy', 'Aspirin');
INSERT INTO shop_orders VALUES (28, 'Sundries', 'Toothbrush');
INSERT INTO shop_orders VALUES (29, 'Sundries', 'Shampoo');
INSERT INTO shop_orders VALUES (30, 'Cereal', 'Bran');
INSERT INTO shop_orders VALUES (31, 'Cereal', 'Corn Flakes');
INSERT INTO shop_orders VALUES (32, 'Cleaners', 'Soap');
INSERT INTO shop_orders VALUES (33, 'Cereal', 'Oatmeal');


There you go! Have fun!

Generating Test Data

Here’s a quick way to generate dummy data for performing tests that require lots of records.

SQL> DROP   TABLE lots_o_data;

Table dropped.

SQL> CREATE TABLE lots_o_data
  2  (	 id	  NUMBER
  3    , name	  VARCHAR2(5)
  4    , category VARCHAR2(26)
  5  );

Table created.

SQL> INSERT INTO lots_o_data VALUES (1, 'Lotso', 'Wow this is a lot of data.');

1 row created.

SQL> INSERT INTO lots_o_data VALUES (2, 'More',  'Bunches and bunches.');

1 row created.


Commit complete.

SQL> INSERT INTO lots_o_data
  2  	 SELECT,, b.category
  3  	 FROM lots_o_data a CROSS JOIN lots_o_data b;

4 rows created.


Commit complete.

SQL> INSERT INTO lots_o_data
  2  	 SELECT,, b.category
  3  	 FROM lots_o_data a CROSS JOIN lots_o_data b;

36 rows created.


Commit complete.

SQL> INSERT INTO lots_o_data
  2  	 SELECT,, b.category
  3  	 FROM lots_o_data a CROSS JOIN lots_o_data b;

1764 rows created.


Commit complete.

SQL> INSERT INTO lots_o_data
  2  	 SELECT,, b.category
  3  	 FROM lots_o_data a CROSS JOIN lots_o_data b;

3261636 rows created.


Commit complete.

  2  FROM   lots_o_data;
  COUNT(*)                                                               ----------                                                             3263442                                                                      

Note the following

  • With only four INSERT … SELECT statements based on a Cartesian product, we were able to generate over 3 million rows of data.
  • A few more INSERT … SELECT statements could easily result in billions of rows of data.

If you need a lot of data for performance testing or capacity testing, this is an easy way to get however much you require.

I have been working on building a data lake environment in anticipation of a series of large data feeds I’m expecting soon and found this to be a useful approach to preparing the environment.

Rumors of SQL’s Death Have Been Greatly Exaggerated

The “NoSQL” revolution is tremendous on many levels.  It’s practical, relevant, and useful.  There are plenty of business situations where an architecture optimized for limited, targeted processing of very large amounts of data simply doesn’t need the overhead of a relational tool and can benefit from specialized functionality optimized for select purposes.

But SQL is still critical to the core operations of any enterprise.  Contrary to what many new-comers might think, “NoSQL” in a given application doesn’t actually mean “No SQL” across the board in the enterprise.  A lot of us who are responsible for enterprise architectures think of the expression “NoSQL” as really saying “Not Only SQL”.  After all – some NoSQL tools are actually built on a SQL platform. Believe it or not, the traditional relational database is still the number one technology for handling the mission critical core services of any professional organization.

However, this truth is apparently lost on some recent arrivals in the world of software development.

About a year ago I was invited to attend a conference in Maryland of one of the leading big data vendors.  Those of us attending were primarily software developers, and we all spoke in terms of various Apache open source projects and their commercial counterparts.  We rarely uttered the word “Java”, yet we were all Java developers of one kind or another.  And those of us in the world of Java development know better about the benefits and relative importance of both NoSQL and SQL platforms.

But I’m describing the attendees.  The conversation I’m about to share was with one of the vendors who was exhibiting at the conference.  (Yes, a commercial vendor was hosting a show at which they invited other commercial vendors, it happens all the time.)  He was trying to sell his company’s product, of course, a product that was complementary to the host company’s core offerings.  His company was one of those now-familiar business models where some developers created a cool open source project that they promoted to the Apache Software Foundation, eventually winning full project status.  With that established, the developers formed a company to begin selling support for the open source ASF product, and also began building a proprietary alternative for sale.  There are dozens of companies doing this nowadays, as we all know – and it’s a great business model, I love it.  More power to them.

So here I was chatting with a sales rep hawking his particular product.  Suddenly he declared the following:

“Well really, relational databases are dead, nobody uses those any more.”

I stopped.  The first visual that popped into my head was of that time as a kid when one of your friends tell you that “nobody” does this or “everybody” does that.  But as you get older, you learn that your fellow six year olds don’t necessarily have a handle on what the whole world is doing.

I’m sure this guy was a smart guy.  And his company was certainly offering a valuable product.  But some perspective was in order.

So I said – really? The RDBMS is dead?  Let me ask you some questions. Don’t you all sell this product?

Him: Yes.

Me: Well – don’t you track sales in a database? Orders, items, linked to the catalog? You probably track orders in an RDBMS.

Him: Well … ok, yes, probably.

Me: And the customers placing those orders? You need data about any number of contact methods, like phone numbers, email addresses, where to ship, when to call, etc., all connected to the orders, sales records, etc.  Probably all structured for mass mailings by correct categories at the right time. Sounds like an RDBMS. Right?

Him: OK, probably, yes.

Me: So … financial data? Employee info and HR compliance, other regulatory issues? Is that going to work in a NoSQL environment?

Him: Good point, probably not.

Me: So the RDBMS is tracking company orders, items, products, customers, employees, suppliers, accounting, finance, human resources, marketing contacts, contracts with events like this conference, and all other data associated with dollars, the law, whatever must be accurately tracked in line with the rest. Even online blog posts are probably running on an RDBMS, right? So that leaves, what, web log files and online comments for marketing trend analysis?

Him: Well no, more than that … but I see your point.

Me;  NoSQL has its place.  So does SQL.  Both are valuable.

Him: Agreed.

One down.

Lots more to go.

– Steve

P.S.  A huge thank you to Steve Feuerstein for encouraging this particular blog post, which is a direct result of a little exchange we had recently on this topic.  Thanks, Steve!

My SQL Expert Book and Practice Exams

I recently read in an online forum the online bonus exam for my book, OCA Oracle Database SQL Certified Expert Exam Guide (Exam 1Z0-047), is “no longer available” at the Oracle Press website.  I’m inquiring into this to determine if its true, and if so, to see if we can get it restored.  If you’ve had a similar experience, please email me to let me know.

In the meantime:

  • I have created additional practice exams here:  They are available at no charge, and only require that you registered an email address so you can take the exams in multiple login sessions if you wish.
  • I am currently working on a revision to my book, to upgrade it for Oracle 12c.  Lots of new feature we’re adding, and I’ll blog about those soon.  The second edition will be completed in early 2016.

In the meantime, if you’re looking for a practice exam, please check out!