MacLochlainns Weblog

Michael McLaughlin's Technical Blog

Site Admin

Archive for the ‘MySQL DBA’ tag

Java & MySQL 8.0.19

without comments

It’s the in-between term time and we’re all stuck at home. I decided to update the image for my Fedora 30 virtual machine. I had a work around to the update issue that I had encountered last October in Bug #96969 but it was not required with the current version. However, after updating from MySQL 8.0.17 to MySQL 8.0.19, I found that my Java connection example failed.

The $CLASSPATH value was correct:

/usr/share/java/mysql-connector-java.jar:.

The first error that I got was the my reference to MySQL JDBC driver was incorrect. The error message is quite clear:

Loading class `com.mysql.jdbc.Driver'. This is deprecated. The new driver class is `com.mysql.cj.jdbc.Driver'. The driver is automatically registered via the SPI and manual loading of the driver class is generally unnecessary.
Cannot connect to database server:
The server time zone value 'MDT' is unrecognized or represents more than one time zone. You must configure either the server or JDBC driver (via the 'serverTimezone' configuration property) to use a more specifc time zone value if you want to utilize time zone support.

I changed the MySQL Driver reference as instructed by the error message:

29
30
31
      // Create instance of MySQLDriver.
      Class.forName ("com.mysql.cj.jdbc.Driver").newInstance();
      conn = DriverManager.getConnection (url, username, password);

After the change, I got the following error while retesting my little MySQL Java driver connection test program. Initially, I thought this required a change in the Java environment but that wasn’t it.

Cannot connect to database server:
The server time zone value 'MDT' is unrecognized or represents more than one time zone. You must configure either the server or JDBC driver (via the 'serverTimezone' configuration property) to use a more specifc time zone value if you want to utilize time zone support.

It required me to add the following line to my /etc/my.cnf configuration file, which synchronizes the database’s timezone with the operating system.

# Synchronize the MySQL clock with the computer system clock.
default-time-zone='+00:00'

Then, running my MySQL Driver connection test program worked like a charm. It returns the following:

Database connection established
MySQLDriver Version [8.0.19]
Database connection terminated

Here’s the MySQL Connector Java code if you’d like to use the MySQLDriver.java file:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
// Import classes.
import java.sql.*;
 
/* You can't include the following on Linux without raising an exception. */
// import com.mysql.jdbc.Driver;
 
public class MySQLDriver {
  public MySQLDriver() {
    /* Declare variables that require explicit assignments because
       they're addressed in the finally block. */
    Connection conn = null;
    Statement stmt = null;
    ResultSet rset = null;
 
    /* Declare other variables. */
    String url;
    String username = "student";
    String password = "student";
    String database = "studentdb";
    String hostname = "localhost";
    String port = "3306";
    String sql;
 
    /* Attempt a connection. */
    try {
      // Set URL.
      url = "jdbc:mysql://" + hostname + ":" + port + "/" + database;
 
      // Create instance of MySQLDriver.
      Class.forName ("com.mysql.cj.jdbc.Driver").newInstance();
      conn = DriverManager.getConnection (url, username, password);
 
      // Query the version of the database.
      sql = "SELECT version()";
      stmt = conn.createStatement();
      rset = stmt.executeQuery(sql);
 
      System.out.println ("Database connection established");
 
      // Read row returns for one column.
      while (rset.next()) {
        System.out.println("MySQLDriver Version [" + rset.getString(1) + "]"); }
 
    }
    catch (SQLException e) {
      System.err.println ("Cannot connect to database server:");
      System.out.println(e.getMessage());
    }
    catch (ClassNotFoundException e) {
      System.err.println ("Cannot find MySQL driver class:");
      System.out.println(e.getMessage());
    }
    catch (InstantiationException e) {
      System.err.println ("Cannot instantiate class:");
      System.out.println(e.getMessage());
    }
    catch (IllegalAccessException e) {
      System.err.println ("Illegal access exception:");
      System.out.println(e.getMessage());
    }
    finally {
      if (conn != null) {
        try {
          rset.close();
          stmt.close();
          conn.close();
          System.out.println ("Database connection terminated");
        }
        catch (Exception e) { /* ignore close errors */ }
      }
    }
  }
  /* Unit test. */
  public static void main(String args[]) {
    new MySQLDriver();
  }
}

As always, I hope this helps those who encounter similar problems.

Written by maclochlainn

April 10th, 2020 at 12:42 pm

Python MySQL Query

without comments

Somebody asked me how to expand a prior example with the static variables so that it took arguments at the command line for the variables. This example uses Python 3 new features in the datetime package.

There’s a small trick converting the string arguments to date data types. Here’s a quick example that shows you how to convert the argument list into individual date data type variables:

#!/usr/bin/python3
 
# include standard modules
import sys
from datetime import datetime
 
# Capture argument list.
fullCmdArguments = sys.argv
 
# Assignable variables.
beginDate = ""
endDate = ""
 
# Assign argument list to variable.
argumentList = fullCmdArguments[1:]
 
# Enumerate through the argument list where beginDate precedes endDate as strings.
try:
  for i, s in enumerate(argumentList):
    if (i == 0):
      beginDate = datetime.date(datetime.fromisoformat(s))
    elif (i == 1):
      endDate = datetime.date(datetime.fromisoformat(s))
except ValueError:
  print("One of the first two arguments is not a valid date (YYYY-MM-DD).")
 
# Print the processed values and types.
print("Begin Date: [",beginDate,"][",type(beginDate),"]")
print("End Date:   [",endDate,"][",type(endDate),"]")

Assume you call this arguments.py. Then, you call it with valid conforming date format value like the following command-line example:

./arguments.py 2001-01-01 2003-12-31

It returns the arguments after they have been converted to date data types. The results should look like this:

Begin Date:  1991-01-01 [ <class 'datetime.date'> ]
End Date:    2004-12-31 [ <class 'datetime.date'> ]

The next Python example accepts dynamic arguments at the command line to query the MySQL database:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
#!/usr/bin/python3
 
# Import the library.
import sys
import mysql.connector
from datetime import datetime
from datetime import date
from mysql.connector import errorcode
 
# Capture argument list.
fullCmdArguments = sys.argv
 
# Assignable variables.
start_date = ""
end_date = ""
 
# Assign argument list to variable.
argumentList = fullCmdArguments[1:]
 
#  Check and process argument list.
# ============================================================
#  If there are less than two arguments provide default values.
#  Else enumerate and convert strings to dates.
# ============================================================
if (len(argumentList) < 2):
  # Set a default start date.
  if (isinstance(start_date,str)):
    start_date = date(1980, 1, 1)
 
  # Set the default end date.
  if (isinstance(end_date,str)):
    end_date = datetime.date(datetime.today())
else:
  # Enumerate through the argument list where beginDate precedes endDate as strings.
  try:
    for i, s in enumerate(argumentList):
      if (i == 0):
        start_date = datetime.date(datetime.fromisoformat(s))
      elif (i == 1):
        end_date = datetime.date(datetime.fromisoformat(s))
  except ValueError:
    print("One of the first two arguments is not a valid date (YYYY-MM-DD).")
 
#  Attempt the query.
# ============================================================
#  Use a try-catch block to manage the connection.
# ============================================================
try:
  # Open connection.
  cnx = mysql.connector.connect(user='student', password='student',
                                host='127.0.0.1',
                                database='studentdb')
  # Create cursor.
  cursor = cnx.cursor()
 
  # Set the query statement.
  query = ("SELECT CASE "
           "         WHEN item_subtitle IS NULL THEN item_title "
           "         ELSE CONCAT(item_title,': ',item_subtitle) "
           "         END AS title, "
           "release_date "
           "FROM item "
           "WHERE release_date BETWEEN %s AND %s "
           "ORDER BY item_title")
 
  # Execute cursor.
  cursor.execute(query, (start_date, end_date))
 
  # Display the rows returned by the query.
  for (item_name, release_date) in cursor:
    print("{}, {:%d-%b-%Y}".format(item_name, release_date))
 
#  Handle exception and close connection.
# ============================================================
except mysql.connector.Error as e:
  if e.errno == errorcode.ER_ACCESS_DENIED_ERROR:
    print("Something is wrong with your user name or password")
  elif e.errno == errorcode.ER_BAD_DB_ERROR:
    print("Database does not exist")
  else:
    print("Error code:", e.errno)        # error number
    print("SQLSTATE value:", e.sqlstate) # SQLSTATE value
    print("Error message:", e.msg)       # error message
 
# Close the connection when the try block completes.
finally:
  cnx.close()

You can call the python-mysql-query.py program with the following syntax:

./python-mysql-query.py 2001-01-01 2003-12-31

It returns the films between 1 Jan 2001 and 31 Dec 2003, like this:

Clear and Present Danger: Special Collector's Edition, 06-May-2003
Die Another Day: 2-Disc Ultimate Version, 03-Jun-2003
Die Another Day, 03-Jun-2003
Die Another Day, 03-Jun-2003
Golden Eye, 03-Jun-2003
Golden Eye: Special Edition, 03-Jun-2003
Harry Potter and the Chamber of Secrets, 28-May-2002
Harry Potter and the Chamber of Secrets: Two-Disc Special Edition, 28-May-2002
Harry Potter and the Sorcerer's Stone, 28-May-2002
Harry Potter and the Sorcerer's Stone: Two-Disc Special Edition, 28-May-2002
Harry Potter and the Sorcerer's Stone: Full Screen Edition, 28-May-2002
MarioKart: Double Dash, 17-Nov-2003
Pirates of the Caribbean, 30-Jun-2003
RoboCop, 24-Jul-2003
Splinter Cell: Chaos Theory, 08-Apr-2003
Star Wars II: Attack of the Clones, 16-May-2002
Star Wars II: Attack of the Clones, 16-May-2002
The Chronicles of Narnia: The Lion, the Witch and the Wardrobe, 30-Jun-2003
The Chronicles of Narnia: The Lion, the Witch and the Wardrobe, 16-May-2002

As always, I hope this helps somebody who wants to learn how to use Python with the MySQL database.

Written by maclochlainn

September 6th, 2019 at 10:31 pm

MySQL Python Connector

with one comment

While building my student image on Fedora 30, I installed the MySQL PHP Connector (php-mysqlndrp) but neglected to install the Python Connector. This adds the installation and basic test of the Python Connector to the original blog post.

You use the following command with a wildcard as a privileged user. The wildcard is necessary because you need to load two libraries to support Python 2.7 and 3.7, which are installed on Fedora 30. You also need to be the root user or a user that is found in the sudoer’s list:

yum install -y mysql-connector-python*

Leveraging the MySQL Connector/Python Coding Examples documentation, Section 5.1 Connecting to MySQL Using Connector/Python here’s a test of the connection to MySQL 8.

# Import the library.
import mysql.connector
from mysql.connector import errorcode
 
try:
  # Open connection.
  cnx = mysql.connector.connect(user='student', password='student',
                                host='127.0.0.1',
                                database='studentdb')
 
  # Print the value.
  print("Database connection resolved.")
 
# Handle exception and close connection.
except mysql.connector.Error as e:
  if e.errno == errorcode.ER_ACCESS_DENIED_ERROR:
    print("Something is wrong with your user name or password")
  elif e.errno == errorcode.ER_BAD_DB_ERROR:
    print("Database does not exist")
  else:
    print(e)
 
# Close the connection when the try block completes.
else:
  cnx.close()

Leveraging the MySQL Connector/Python Coding Examples documentation, Section 5.4 Querying Data Using Connector/Python here’s a test of the connection to MySQL 8.

# Import the library.
import datetime
import mysql.connector
from mysql.connector import errorcode
 
try:
  # Open connection.
  cnx = mysql.connector.connect(user='student', password='student',
                                host='127.0.0.1',
                                database='studentdb')
  # Create cursor.
  cursor = cnx.cursor()
 
  # Set the query statement.
  query = ("SELECT "
           "CASE "
           "  WHEN item_subtitle IS NULL THEN item_title "
           "  ELSE CONCAT(item_title,': ',item_subtitle) "
           "END AS title, "
           "release_date "
           "FROM item "
           "WHERE release_date BETWEEN %s AND %s "
           "ORDER BY item_title")
 
  # Set the start and end date.
  start_date = datetime.date(1991, 1, 1)
  end_date = datetime.date(2004, 12, 31)
 
  # Execute cursor.
  cursor.execute(query, (start_date, end_date))
 
  # Display the rows returned by the query.
  for (item_name, release_date) in cursor:
    print("{}, {:%d %b %Y}".format(item_name, release_date))
 
  # Close cursor.
  cursor.close()
 
# ------------------------------------------------------------
# Handle exception and close connection.
except mysql.connector.Error as e:
  if e.errno == errorcode.ER_ACCESS_DENIED_ERROR:
    print("Something is wrong with your user name or password")
  elif e.errno == errorcode.ER_BAD_DB_ERROR:
    print("Database does not exist")
  else:
    print "Error code:", e.errno        # error number
    print "SQLSTATE value:", e.sqlstate # SQLSTATE value
    print "Error message:", e.msg       # error message
 
# Close the connection when the try block completes.
else:
  cnx.close()

If you run the above in Python 2.7 it works fine. It fails to parse successfully in Python 3.x because the print() function requires the parentheses all the time. You would need to re-write the except block, like this with the parentheses:

# Handle exception and close connection.
except mysql.connector.Error as e:
  if e.errno == errorcode.ER_ACCESS_DENIED_ERROR:
    print("Something is wrong with your user name or password")
  elif e.errno == errorcode.ER_BAD_DB_ERROR:
    print("Database does not exist")
  else:
    print("Error code:", e.errno)        # error number
    print("SQLSTATE value:", e.sqlstate) # SQLSTATE value
    print("Error message:", e.msg)       # error message

While it works without the parentheses in Python 2.7, it also works with the parentheses. That means the best practice is to write cross compatible code by always using the parentheses with the print() function.

As always, I hope this helps somebody.j

Written by maclochlainn

August 21st, 2019 at 1:44 am

Apache on Fedora 30

with one comment

There was an option during the Fedora 30 Workstation installation to add the Apache Web Server, but you need to set it to start automatically. Unfortunately, there was no option to install PHP, which I thought odd because of how many web developers learn the trade first on PHP with a LAMP (Linux, Apache, MySQL, Perl/PHP/Python) stack. You see how to fix that shortcoming in this post and how to install and test PHP, mysqli, and pdo to support MySQL 8.

Before you do that make sure you install MySQL 8. You can find my prior blog post on that here.

You set Apache to start automatically, on the next boot of the operating system, with the following command:

chkconfig httpd on

It creates a symbolic link:

Created symlink /etc/systemd/system/multi-user.target.wants/httpd.service → /usr/lib/systemd/system/httpd.service.

However, that command only starts the Apache server the next time you boot the server. You use the following command as the root user to start the Apache server:

apachectl start

You can verify the installation with the following command as the root user:

ps -ef | grep httpd | grep -v grep

It should return:

root      5433     1  0 17:03 ?        00:00:00 /usr/sbin/httpd -DFOREGROUND
apache    5434  5433  0 17:03 ?        00:00:00 /usr/sbin/httpd -DFOREGROUND
apache    5435  5433  0 17:03 ?        00:00:00 /usr/sbin/httpd -DFOREGROUND
apache    5436  5433  0 17:03 ?        00:00:00 /usr/sbin/httpd -DFOREGROUND
apache    5437  5433  0 17:03 ?        00:00:00 /usr/sbin/httpd -DFOREGROUND
apache    5438  5433  0 17:03 ?        00:00:00 /usr/sbin/httpd -DFOREGROUND
apache    5442  5433  0 17:03 ?        00:00:00 /usr/sbin/httpd -DFOREGROUND

and, then verify the listening port with the following command as the root user:

netstat -tulpn | grep :80

It should return the following when both the Apache server is listening on port 80 and the Oracle multi-protocol server is listening on port 8080:

tcp6       0      0 :::80                   :::*                    LISTEN      119810/httpd        
tcp6       0      0 :::8080                 :::*                    LISTEN      1403/tnslsnr

You can also enter the following URL in the browser to see the Apache Test Page:

http://localhost

It should display the test page, like this:

You can also create a hello.htm file in the /var/www/html directory to test the ability to read an HTML file. I would suggest the traditional hello.htm file:

<html>
<body>
Hello World!
</body>
</html>

You can call it by using this URL in the browser:

http://localhost/hello.htm

It should display the test page, like this:

Now, let’s install PHP. You use the following command as a privileged user, which is one found in the sudoer’s list:

yum install -y php

Before you test the installation of PHP in a browser, you must restart the Apache HTTP Server. You can do that with the following command as a privileged user:

sudo apachectl restart

After verifying the connection, you can test it by creating the traditional info.php program file in the /var/www/http directory. The file should contain the following:

1
2
3
<?php
  phpinfo();
?>

It should display the PHP Version 7.3.8 web page, which ships with Fedora 30:

The next step shows you how to install mysqli and pdo with the yum utility. While it’s unnecessary to check for the older mysql library (truly deprecated), its good practice to know how to check for a conflicting library before installing a new one. Also, I’d prefer newbies get exposed to using the yum utility’s shell environment.

You start the yum shell, as follows:

yum shell

With the yum shell, you would remove a mysql package with the following command:

> remove php-mysql

The command will remove the package or tell you that there is no package to remove. Next, you install the php-mysqli package with this command:

install php-mysqli

You will then be prompted to confirm the installation of the php-mysqli library. Finally, you exit the yum shell with this command:

> quit

If you want to see the whole interactive shell, click on the link below.

You need to restart the Apache HTTP listener for these changes to take place, which you do with the same command as shown earlier:

sudo apachectl restart

I wrote the mysqli_check.php script to verify installation of both the mysqli and pdo libraries. The full code should be put in a mysqli_check.php file in the /var/www/html directory for testing.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
<html>
<header>
<title>Static Query Object Sample</title>
<style type="text/css">
  /* HTML element styles. */
  table {background:white;border-style:solid;border-width:3px;border-color:black;border-collapse:collapse;}
  th {text-align:center;font-style:bold;background:lightgray;border:solid 1px gray;}
  td {border:solid 1px gray;}
 
  /* Class tag element styles. */
  .ID {min-width:50px;text-align:right;}
  .Label {min-width:200px;text-align:left;}
</style>
</header>
<body>
<?php
  if (!function_exists('mysqli_init') && !extension_loaded('mysqli')) {
    print 'mysqli not installed.'; }
  else {
    print 'mysqli installed.'; }
  if (!function_exists('pdo_init') && !extension_loaded('pdo')) {
    print '<p>pdo not installed.</p>'; }
  else {
    print '<p>pdo installed.</p>'; }
?>
</script>
</body>
</html>

You can test it with the following URL from the local browser:

http://localhost/mysqli_check.php

It should print the following to the web page when you’ve successfully install the mysqli and pdo libraries:

mysqli installed.
pdo installed.

If you plan to use PHP to display and render graphics, you need to install php-gd library. You can do that with the yum utility and this prior blog post explains it. Don’t forget to restart the Apache HTTP Server after you add the php-gd library.

For example, one of my sample PHP programs loads a PNG image into a BLOB column as raw binary text. Then, the program reads it and renders it with PHP to produce the following web page.

As always, I hope this helps those looking for a complete solution without cost.

Written by maclochlainn

August 16th, 2019 at 12:26 pm

MySQL on Fedora 30

with one comment

While updating my class image to Fedora 30, I noticed that it installed the Akonadi Server. The documentation on the Akonadi server lacked some straightforward documentation. It also offered a bundled set of software that limited how to approach MySQL development.

So, I removed all those packages with the following syntax:

dnf remove `rpm -qa | grep akonadi`

After removing those Akonadi packages, I installed the MySQL Community Edition from the Fedora repo with this syntax:

yum install -y community-mysql*

Having installed MySQL Community Edition, I wanted to start the mysql service with this command:

sudo service mysqld start

Unfortunately, the service utility wasn’t installed. That surprised me. While I could have run this command:

systemctl start mysqld.service

A better solution was to install any missing code components. I determined that the service utility is part of the initscripts package; and I installed it with the following command:

sudo yum install -y initscripts

Then, I ran the mysql_secure_installation script to secure the installation:

mysql_secure_installation

The script set the root user’s password, remove the anonymous user, disallow remote root login, and remove the test databases. Then, I verified connecting to the MySQL database with the following syntax:

mysql -uroot -ppassword

I enabled the MySQL Service to start with each reboot of the Fedora instance. I used the following command:

systemctl enable mysqld.service

It creates the following link:

ln -s '/etc/systemd/system/multi-user.target.wants/mysqld.service' '/usr/lib/systemd/system/mysqld.service'

The next step requires setting up a sample studentdb database. The syntax has changed from prior releases. Here are the three steps:

  1. Create the studentdb database with the following command as the MySQL root user:

    mysql> CREATE DATABASE studentdb;
  2. Grant the root user the privilege to grant to others, which root does not have by default. You use the following syntax as the MySQL root user:

    mysql> GRANT ALL ON *.* TO 'root'@'localhost';
  3. Create the user with a clear English password and grant the user student full privileges on the studentdb database:

    mysql> CREATE USER 'student'@'localhost' IDENTIFIED WITH mysql_native_password BY 'student';
    mysql> GRANT ALL ON studentdb.* TO 'student'@'localhost';

If you fail to specify mysql_native_password when creating the user and use the older syntax like the following example:

mysql> CREATE USER 'student'@'localhost' IDENTIFIED BY 'student';
mysql> GRANT ALL ON studentdb.* TO 'student'@'localhost';

The GRANT command will raise the following error:

ERROR 1410 (42000): You are not allowed to create a user with GRANT

Written by maclochlainn

August 16th, 2019 at 1:02 am

Posted in MySQL,MySQL 8

Tagged with ,

MySQL 5.7.* and mysqli

without comments

After installing MySQL 5.7.22 and PHP 7.1.17 on Fedora 27, you need to install the mysqli library. You need to verify if the mysqli library is installed. You can do that with the following mysqli_check.php program:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
<html>
<header>
<title>Check mysqli Install</title>
</header>
<body>
<?php
  if (!function_exists('mysqli_init') && !extension_loaded('mysqli')) {
    print 'mysqli not installed.'; }
  else {
    print 'mysqli installed.'; }
?>
</script>
</body>
</html>

You test preceding PHP program with the following URL in a browser:

http://localhost/mysqli_check.php

If the mysqli program isn’t installed, you can install it as follows by opening the yum interactive shell:

[root@localhost html]# yum shell
Last metadata expiration check: 1:26:46 ago on Wed 22 Aug 2018 08:05:50 PM MDT.
> remove php-mysql
No match for argument: php-mysql
Error: No packages marked for removal.
> install php-mysqlnd
> run
================================================================================================
 Package                 Arch               Version                   Repository           Size
================================================================================================
Installing:
 php-mysqlnd             x86_64             7.1.20-1.fc27             updates             246 k
Upgrading:
 php                     x86_64             7.1.20-1.fc27             updates             2.8 M
 php-cli                 x86_64             7.1.20-1.fc27             updates             4.2 M
 php-common              x86_64             7.1.20-1.fc27             updates             1.0 M
 php-fpm                 x86_64             7.1.20-1.fc27             updates             1.5 M
 php-json                x86_64             7.1.20-1.fc27             updates              73 k
 php-pdo                 x86_64             7.1.20-1.fc27             updates             138 k
 php-pgsql               x86_64             7.1.20-1.fc27             updates             135 k
 
Transaction Summary
================================================================================================
Install  1 Package
Upgrade  7 Packages
 
Total download size: 10 M
Is this ok [y/N]: y

After you type y and the return key, you should see a detailed log of the installation. Click the link below to see the yum installation log detail.

After you install the mysqli library, you exit the yum interactive shell with the quit command as shown:

> quit
Leaving Shell
The downloaded packages were saved in cache until the next successful transaction.
You can remove cached packages by executing 'dnf clean packages'.

You can now retest by re-running the mysqli_check.php program with the following URL:

http://localhost/mysqli_check.php

Image processing is not generally installed by default. You should use the following yum command to install the PHP Image processing library:

yum install -y php-gd

Or, you can use dnf (Dandified yum), like:

dnf install -y php-gd

Click the link below to see the yum installation log detail.

If you encounter an error trying to render an image like this:

Call to undefined function imagecreatefromstring() in ...

The php-gd package is not enabled. You can verify the contents of the php-gd package with the following rpm command on Fedora or CentOS:

rpm -ql php-gd

On PHP 7.1, it should return:

/etc/php-zts.d/20-gd.ini
/etc/php.d/20-gd.ini
/usr/lib/.build-id
/usr/lib/.build-id/50
/usr/lib/.build-id/50/11f0ec947836c6b0d325084841c05255197131
/usr/lib/.build-id/b0/10bf6f48ca6c0710dcc5777c07059b2acece77
/usr/lib64/php-zts/modules/gd.so
/usr/lib64/php/modules/gd.so

Then, you might choose to follow some obsolete note from ten or more years ago to include gd.so in your /etc/php.ini file. That’s not necessary.

The most common reason for incurring this error is tied to migrating old PHP 5 code forward. Sometimes folks used logic like the following to print a Portable Network Graphics (png) file stored natively in a MySQL BLOB column:

  header('Content-Type: image/x-png');
  imagepng(imagecreatefromstring($image));

If it was stored as a Portable Network Graphics (png) file, all you needed was:

  header('Content-Type: image/x-png');
  print $image;

As always, I hope this helps those looking for a solution.

Written by maclochlainn

August 23rd, 2018 at 11:47 am

MySQL on Fedora 27

without comments

While updating my class image to Fedora 27, I noticed that it installed the Akonadi Server. The documentation on the Akonadi server lacked some straightforward documentation. It also offered a bundled set of software that limited how to approach MySQL development.

So, I removed all those packages with the following syntax:

dnf remove `rpm -qa | grep akonadi`

After removing those Akonadi packages, I installed the MySQL Community Edition from the Fedora repo with this syntax:

yum install -y community-mysql*

Having installed MySQL Community Edition, I started the service with this command:

sudo service mysql start

Then, I ran the mysql_secure_installation script to secure the installation:

mysql_secure_installation

The script set the root user’s password, remove the anonymous user, disallow remote root login, and remove the test databases. Then, I verified connecting to the MySQL database with the following syntax:

mysql -uroot -ppassword

I enabled the MySQL Service to start with each reboot of the Fedora instance. I used the following command:

systemctl enable mysqld.service

It creates the following link:

ln -s '/etc/systemd/system/multi-user.target.wants/mysqld.service' '/usr/lib/systemd/system/mysqld.service'

Now, I need to install and configure Apache, PHP, and upgrade Oracle Database 11g XE’s APEX 4 to 5.

Written by maclochlainn

May 16th, 2018 at 2:13 am

PostgreSQL Identity Columns

without comments

It’s interesting to see the way different databases implement automatic numbering. Oracle Database 12c is the closest to PostgreSQL in some significant ways. However, its probably more accurate to say Oracle Database 12c copied PostgreSQL’s implementation. At least, that’s my conjecture because Oracle added a way to reset the START WITH value of the indirect sequence. However, I prefer the MySQL approach because the automatic numbering sequence is a property of the table and a simple clause of the CREATE TABLE statement.

Both PostgreSQL and Oracle Database 12c implement automatic numbering as indirect sequences. Indirect sequences are those created by a table when you designate a column as an identity column in Oracle or as a serial column in PostgreSQL. The difference is that PostgreSQL doesn’t provide a syntax version inside the CREATE TABLE semantic.

MySQL provides such syntax. You set an auto numbering column in MySQL by appending the AUTO_INCREMENT clause to the table creation statement when you want it to start with a number other than 1, like this:

CREATE TABLE auto
( id           INT UNSIGNED PRIMARY KEY AUTO_INCREMENT
, text_field   VARCHAR(30)  NOT NULL
) ENGINE=InnoDB AUTO_INCREMENT=1001 DEFAULT CHARSET=utf8;

Oracle disallows you to changing a sequence created as a background activity of the CREATE TABLE statement; and Oracle disallows you dropping an indirect sequence without changing the table that created it, which is exactly how they handle indexes created for unique constraints. Unfortunately, Oracle also disallows altering the START WITH value of any sequence.

If you want to change the START WITH value on an Oracle Database 12c indirect sequence, you must export the table, drop the table, and recreate the table with a new START WITH value before importing the data back into the table. The syntax for setting an IDENTITY column value higher than 1 is:

CREATE TABLE auto
( auto_id     NUMBER GENERATED BY DEFAULT AS IDENTITY (START WITH 1001)
, text_field  VARCHAR2(30)
, CONSTRAINT  auto_pk PRIMARY KEY (auto_id));

You can only create a PostgreSQL table with automatic numbering by using the SERIAL data type, which always sets the initial value to 1. You can reset the SERIAL sequence value in PostgreSQL with the ALTER statement. Unlike Oracle Database 12c, PostgreSQL does let you modify the START WITH value of any sequence. The trick is understanding how to find the sequence name. The name is always the combination of the table name, an underscore, an id string, an underscore, and a seq string. This behavior makes a great case for choosing id as the name of any auto numbering columns in a table.

CREATE TABLE auto
( id          SERIAL      CONSTRAINT auto_pk PRIMARY KEY
, text_field  VARCHAR(30));
 
ALTER SEQUENCE auto_id_seq RESTART WITH 1001;

You can see the table and assigned sequence with the following command in PostgreSQL:

\d+ auto

It should display:

                                                      Table "public.auto"
   Column   |         Type          |                     Modifiers                     | Storage  | Stats target | Description 
------------+-----------------------+---------------------------------------------------+----------+--------------+-------------
 id         | integer               | not null default nextval('auto_id_seq'::regclass) | plain    |              | 
 text_field | character varying(30) |                                                   | extended |              | 
Indexes:
    "auto_pk" PRIMARY KEY, btree (id)
Has OIDs: no

As always, I hope this helps those trying to sort through how to start identity columns above the initial value of 1.

Written by maclochlainn

August 4th, 2017 at 12:52 am

Fedora LAMP Steps

without comments

I posted earlier in the year how to configure a Fedora instance to test PHP code on a local VM. However, I’ve got a few questions on how to find those posts. Here’s a consolidation with links on those steps:

  1. Go to this blog post and install the httpd and php libraries with the yum installer.
  2. In the same blog post as step 1 (you can put the sample PHP code into the /var/www/html directory for testing), connect to the yum shell and remove the php-mysql library and then install the mysqlnd library.
  3. Go to this blog post and install the php-gd libraries, which enable you to render PNG images stored as binary streams in MySQL.

As always, I hope that helps.

Written by maclochlainn

December 9th, 2015 at 9:44 am

Create MySQL Index

without comments

Indexes are separate data structures that provide alternate pathways to finding data. They can and do generally speed up the processing of queries and other DML commands, like the INSERT, UPDATE, REPLACE INTO, and DELETE statements. Indexes are also called fast access paths.

In the scope of the InnoDB Database Engine, the MySQL database maintains the integrity of indexes after you create them. The upside of indexes is that they can improve SQL statement performance. The downside is that they impose overhead on every INSERT, UPDATE, REPLACE INTO, and DELETE statement, because the database maintains them by inserting, updating, or deleting items for each related change in the tables that the indexes support.

Indexes have two key properties—usability and visibility. Indexes are both usable and visible by default. That means they are visible to the MySQL cost-based optimizer and usable when statements run against the tables they support.

You have the ability to make any index invisible, in which case queries and DML statements won’t use the index because they won’t see it. However, the cost-based optimizer still sees the index and maintains it with any DML statement change. That means making an index invisible isn’t quite like making the index unusable or like dropping it temporarily. An invisible index becomes overhead and thus is typically a short-term solution to run a resource-intensive statement that behaves better without the index while avoiding the cost of rebuilding it after the statement runs.

It is also possible to make an index unusable, in which case it stops collecting information and becomes obsolete and the database drops its index segment. You rebuild the index when you change it back to a usable index.

Indexes work on the principal of a key. A key is typically a set of columns or expressions on which you can build an index, but it’s possible that a key can be a single column. An index based on a set of columns is a composite, or concatenated, index.

Indexes can be unique or non-unique. You create a unique index anytime you constrain a column by assigning a primary key or unique constraint, but they’re indirect indexes. You create a direct unique index on a single column with the following syntax against two non-unique columns:

1
2
CREATE INDEX common_lookup_u1
ON common_lookup (common_lookup_table) USING BTREE;

You could convert this to a non-unique index on two columns by using this syntax:

1
2
CREATE INDEX common_lookup_u1
ON common_lookup (common_lookup_table, common_lookup_column) USING BTREE;

Making the index unique is straightforward;, you only need to add a UNIQUE key wordk to the CREATE INDEX statement, like

1
2
3
4
CREATE UNIQUE INDEX common_lookup_u1
ON common_lookup ( common_lookup_table
                 , common_lookup_column
                 , common_lookup_type) USING BTREE;

Most indexes use a B-tree (balanced tree). A B-tree is composed of three types of blocks—a root branch block for searching next-level blocks, branch blocks for searching other branch blocks, or and leaf blocks that store pointers to row values. B-trees are balanced because all leaf-blocks are at the same level, which means the length of search is the same to any element in the tree. All branch blocks store the minimum key prefix required to make branching decisions through the B-tree.

Written by maclochlainn

September 29th, 2015 at 6:41 pm