DBD::mysql 4.0.23 and MySQLBind released
For the sake of getting things done and not having projects succumb to bit-rot (in the case of MySQLBind), today I have released DBD::mysql 4.0.23 and MySQLBind with prepared-statement support.
DBD::mysql is the Perl driver for MySQL. I have been maintaining this project now for almost 10 years. I work on a lot of different projects and various languages, but Perl is very dear to me and so many things taken for granted that we use in the technical world use it and it Just Works (TM). MySQL is of course dear to me as well, and Perl talking to MySQL is of utmost importance, hence why I love maintaining this driver. I know the code in my sleep as well! This version includes various changes from contributors in the Open Source community who I am very grateful to. Please see the change log for details!
Also-- this is very important -- I want to thank Sveta Smirnova (Спасибо Света!) as she took painstaking effort to transfer all the DBD::mysql bugs from Oracle's bug system for MySQL to http://rt.cpan.org
. I really appreciate this, even though there are a mountain of bugs to fix (some fixed already). I have had offers to help with that I intend to accept! Thank you to MySQL AB, Sun, Oracle, et al, for having helped with bugs throughout the years.
The code is at:https://github.com/CaptTofu/DBD-mysql
As well as:http://search.cpan.org/~capttofu/DBD-mysql-4.023/
(allow time for mirrors to update)
This is a project developed by Robert Prouse, ~2004. It essentially makes it possible to use MySQL as a store for DNS records for Bind 9 vs. files hence only requiring a Bind reload (using rndc) when a zone is added or deleted versus having to reload for any zone record change when using files. Over the last several months, I have been working on DNSaaS at HP. One of the early proto-types I was working on with Kiall I had created an agent to talk to MySQLBind, which I initially thought we'd use, though we went instead with PowerDNS (great choice!). The code for MySQLBind works quite well. Additionally, I converted all database API calls to utilize the prepared statement API as well as updated the driver to use a single database table whereas before it used a table-per-zone which I felt was untenable for management and scaling. The single table I have provided SQL creation files for and show how you can use partitioning to make it possible to scale out -- if you need to and have so many domains you need to worry about this good problem to have! Also, I added a column "tenant_id". This is very specific to Moniker (https://wiki.openstack.org/wiki/Moniker
), but ultimately you can simply use some unique per-domain identifier which is the key used for partitioning. The code is at https://github.com/CaptTofu/mysql-bind