Mysql for excel for mac

Ever tried to access a mySQL database within Excel on macOS using the ODBC driver (64bit)? – I tried and failed. Here is my way out of the mess.

Using a Windows environment it is so pretty simple: Download ODBC driver for Windows from the mySQL Webpage, install it, open the system control -> ODBC configuration, add a DSN using input fields, test it. Next start Excel, external datasource, select DSN and voila: full access to the database.

I’ll spare you all the details of failure and just list a few of the error messages on the way to success:

The installation failed. The Installer encountered an error the caused the installation to fail. Contact the software manufacturer for assistance.

Error message if you want to install the MySQL Connector/Odbc

[iODBC][Driver Manager]Specified driver could not be loadedError message in Excel, if you want to address the database

And now, how it worked for me

iODBC Environment

First of all you have to install the iODBC (3.52.12 or later) environment on the OS X system before your can install the Connector/ODBC. You’ll find the installation packet here:http://www.iodbc.org/dataspace/doc/iodbc/wiki/iodbcWiki/Downloads. Just download and install it. You’ll find the iODBC Administrator in “Programs -> iODBC”

ODBC/Connector mySQL

Before using the iODBC Administrator download an install the ODBC/Connector for MySQL. You’ll find the installation packet here: https://dev.mysql.com/downloads/connector/. Now that the iODBC package is installed, this installation works without the error message “The installation failed”.

Moving the ODBC/Connector

And now a step that makes the use with Excel possible. The Microsoft Office solution under macOS is a Sandboxed application and doesn’t have access to the default installation path of the ODBC/Connector (/usr/local/mysql-connector-odbc …..). You’ll have to move the installation to /Library/ODBC … And you have to do some more modifications. I found the following script (Sam/samsgit on guthub.com) doing this in a perfect way:

#!/bin/bash
 # https://github.com/openlink/iODBC/issues/29
# https://bugs.mysql.com/bug.php?id=89931
 base_src_dir="/usr/local"
mysql_odbc_name=$(ls "$base_src_dir" | grep -m 1 "mysql-connector-odbc")
odbc_dir="/Library/ODBC"



  


                       
  



 src="$base_src_dir/$mysql_odbc_name/lib"
dst="$odbc_dir/$mysql_odbc_name/lib"
 echo "creating '$dst'"
sudo mkdir -p "$dst"
 echo "copying '$src' to '$dst'"
sudo cp -af "$src/." "$dst"
  odbc_ini_path="$odbc_dir/odbc.ini"
odbc_ini_bak_path="$odbc_ini_path.bak"
 odbcinst_ini_path="$odbc_dir/odbcinst.ini"
odbcinst_ini_bak_path="$odbcinst_ini_path.bak"
 echo "backing up '$odbc_ini_path' to '$odbc_ini_bak_path'"
sudo cp -f "$odbc_ini_path" "$odbc_ini_bak_path"
 echo "backing up '$odbcinst_ini_path' to '$odbcinst_ini_bak_path'"
sudo cp -f "$odbcinst_ini_path" "$odbcinst_ini_bak_path"
 # https://stackoverflow.com/a/29626460
function replace {
  sudo sed -i '' "s/$(sed 's/[^^]/[&]/g; s/^/\^/g' <<< "$1")/$(sed 's/[&/]/\&/g' <<< "$2")/g" "$3"
}



  


                       
  



 ansi_driver=$(ls "$dst" | grep -m 1 "^lib.*a.so$")
unicode_driver=$(ls "$dst" | grep -m 1 "^lib.*w.so$")
 old_ansi_path="$src/$ansi_driver"
new_ansi_path="$dst/$ansi_driver"
 old_unicode_path="$src/$unicode_driver"
new_unicode_path="$dst/$unicode_driver"
 echo "updating '$old_ansi_path' to '$new_ansi_path' in '$odbc_ini_path'"
replace "$old_ansi_path" "$new_ansi_path" "$odbc_ini_path"
 echo "updating '$old_ansi_path' to '$new_ansi_path' in '$odbcinst_ini_path'"
replace "$old_ansi_path" "$new_ansi_path" "$odbcinst_ini_path"
 echo "updating '$old_unicode_path' to '$new_unicode_path' in '$odbc_ini_path'"
replace "$old_unicode_path" "$new_unicode_path" "$odbc_ini_path"
 echo "updating '$old_unicode_path' to '$new_unicode_path' in '$odbcinst_ini_path'"
replace "$old_unicode_path" "$new_unicode_path" "$odbcinst_ini_path"

Just open a terminal window, create a new file with an editor (eg. vi <script name>) and paste the script above. After saving the file make the script executable for your OS (eg. chmod a+x <script name>). Last but not least, run the script (eg. “./<script name>).

Configure the ODBC/Connector

Now you can configure the DSN in the iODBC Administrator. Just add an User DSN:

And it works

Now you can start Excel and navigate to “data” -> “new query” -> “Database” and select your DSN entry 🙂

Update 6/2021

I have reproduced this setup today (June 2021) with the current drivers and Microsoft program versions. Unfortunately, Microsoft has changed the way the ODBC drivers are integrated in the current versions. All tests on the ODBC driver run positively, but unfortunately, it does not work in Excel. 

Microsoft now refers to a paid ODBC plugin from a third party. Apparently they want to push this and have in this course prevented the alternative 🙁

I have numerous Excel workbooks that I use to keep track of HR-related information. In order to minimize data duplication, I have consolidated these records into as few files as possible, but some of the core information such as Names, Employee ID numbers, and Job Titles are necessarily duplicated between files.

Wouldn’t it be great if the main information could just be updated in one place and then propagate across various systems and files? I’m working on porting all of my Excel files into a web-based system, but it’s really a side-project, so in the meantime I wanted a solution to reduce this duplication, and I think I may have found a realistic solution.

Why not keep the main employee information, which is currently duplicated across several Excel files, in a MySQL database and then link this into each of my Excel workbooks so that any update to these data will automatically update in all the files. As an added bonus, this database could later form the core of the web-based system, allowing me to implement features incrementally.

I already have a MySQL server on the local network that synchronizes attendance data from our fingerprint attendance device. So I’m going to use the employees table in that database as the source for my Excel files to get their employee data.

Here are my notes of how I got it working on Excel 16.32 running on Mac OS 10.15.1 connecting via LAN to MySQL 5.7.28 on Ubuntu 18.04 LTS.

Setup

On MySQL server

By default, MySQL will only accept connections from the local machine, and the default recommended privileges for a new user also include @localhost.

Side note:
You can check what ports and addresses your MySQL is currently listening on with:

$ netstat -tulpen

To make sure you will be able to connect (leaving any firewall configuration changes up to you):

1. Change the default bind-address:

$ sudo nano /etc/mysql/mysql.conf.d/mysqld.cnf

On some versions you may need to use /etc/mysql/conf.d/mysql.cnf or /etc/mysql/my.cnf.

Different versions appear to work differently as well, so either comment out the existing bind-address:

# Instead of skip-networking the default is now to listen only on
# localhost which is more compatible and is not less secure.
# bind-address = 127.0.0.1

or change the existing bind-address line to:

bind-address = 0.0.0.0

Then restart mysqld

$ sudo /etc/init.d/mysql restart

2. Add a new user with the @ set to your computer’s IP address if you have a static IP, or else use a wildcard for computers on your local subnet:

GRANT ALL PRIVILEGES ON *.* to 'remoteuser’@’10.1.1.%' IDENTIFIED BY 'remoteuserspassword' WITH GRANT OPTION;

Obviously, adjust as necessary, and if copying and pasting make sure to double-check that quote marks and dashes etc haven’t been autocorrected to pretty versions.

On MacOS

1. Install ODBC Manager (Freeware alternative to ODBC Administrator)
   http://www.odbcmanager.net/
2. Install the MySQL connector
   https://dev.mysql.com/downloads/connector/odbc/
3. Move the installed connector from local to global library
   https://stackoverflow.com/questions/52896893/macos-connector-mysql-odbc-driver-could-not-be-loaded-in-excel-for-mac-2016 quoted below:
   

   I was able to connect to MySQL with Microsoft Excel Version 16.19 by making the following adjustments:
   1. I copied the driver’s whole directory /mysql-connector-odbc-8.0.12-macos10.13-x86-64bit to /Library/ODBC (why? see https://bugs.mysql.com/bug.php?id=89931, and Is there a way to get ADODB to work with Excel for Mac 2011?) ; basically the reason for that is the driver gets downloaded to /usr/local and Excel, being sandboxed, cannot access that location.
   2. Updated the file odbc.ini, also located in /Library/ODBC, with the driver’s new location.

4. Run ODBC manager and add your remote MySQL database with options SERVER and PORT set appropriately (see the default entries for syntax). Click on Test to make sure the connection works.

Making the Connection

Open Excel and navigate to the Data tab, select New Database Query, From Database.

In the iODBC Data Source Chooser window that opens, select the new ODBC Data Source you just added in ODBC manager and click OK. Enter your new MySQL username and password when prompted.

Enter a single-line query and click on Run. If no errors occur (and reasonable results appear in the preview pane) then send the results back to Excel by clicking on Return Data. Choose whether to add the results to the current worksheet, a new worksheet, or a pivot table. You can also click on Properties to choose when and how frequently the data will be refreshed. Note that if the database is on an intranet server and you open the Excel sheet from home, for example, the file will still open but it will just use the version of the data from when it was last opened.

This blog post shows you how to configure Excel 2011 on Mac OS X to query a MySQL database natively installed on the same Mac. If you need installation instructions for the MySQL database natively on Mac OS X, use this post. The configuration is only required the first time and then you may reuse the connection later.

Before you can begin these steps inside Excel 2011, you need to download and install OpenLink Software’s ODBC Driver. I’ve posted instructions in this other blog page because all the screen shots make page loading a problem. You open a new Workbook and click on the Data tab. Then, on the Database icon in the Data ribbon. It lets you import data from the database.

The configuration steps are as follows:

1. This first step prompt you to install Rosetta which allows native Power PC applications to run your Intel-based Mac OS X. It will eventually no longer be supported with the release of Lion (at least that’s the rumor at the time of writing). Click the Install button to get this working.

1. The install took a minute on my Mac Pro, so I figured you should see the progress bar in case it takes a half minute or so.

1. After the completion, you now configure a Data Source Name. This type of configuration is provided for by components on the Windows OS (see this post for an example), which don’t exist on Mac OS X. It appears that Microsoft didn’t want to provide those components, which are a dependency for Excel 2011, and that’s why you need OpenLink Software, specifically the iODBC Data Source Connector. You should note that it looks virtually the same as the Windows OS component. Click the Add button to proceed.

1. The next screen lets you choose a driver from those installed on the Mac OS X system. If you went to the bother to install the free MySQL Connector/ODBC (instructions here), it won’t work because the Microsoft Component requires Rosetta to work. It appears that Microsoft Query isn’t a native Intel-port but rather a Power PC port. Choose one of the OpenLinnk drivers and click the Finish button to continue.

1. The next step requires that you configure the Data Source Name (DSN), provide a hostname and listener port for MySQL (the default port is 3306). After you enter these values, click the Connection tab to the right of the Data Source tab in the dialog box.

1. This step requires that you set the user name, password, and database. The ability to pick the database in the iODBC Data Source Connector is the result of the MySQL Lite tool. It requires that you’ve configured the database first. After entering the data, click the Options tab in the dialog to set the next set of values.

1. This step doesn’t initially let you do much but afterward you can set the character set. Click the Preference tab to proceed with the configuration.

1. There’s nothing you need do here, but I’d suggest in a real situation that you click the Always include VIEWS in table list check box before you click the Finish button.

1. The Data Source Name setup is complete, the next screen lets you launch a connection to the MySQL database. Click the OK button to complete the configuration.

1. At this point, you repeat the Step #1. It launches the iODBC Data Source Chooser dialog. Click the Data Source Name and then the Test button.

1. The test launches a dialog to test the connection, as shown below. Enter the Password in the dialog.

1. If you’ve configured everything correctly, you’ll see the following confirmation dialog. Click the OK button.

Hope this helps you if you’re looking to connect Excel 2011 to query MySQL.