I'm extremely happy to announce that the MariaDB project has integrated FederatedX into the main tree and has replaced the old, un-maintained Federated storage engine! This means that from now on, I will target any FederatedX changes and enhancements for MariaDB since it will be easier than having to try to make it run as a pluggable storage engine. This also means I will have another reason why I should keep this project moving forward. I will still provide the ability for anyone who wants to use it as a pluggable engine (not that difficult) so this can be loaded with MySQL as well.
Also changed: I changed to using the BSD license for FederatedX.
For Drizzle, I will need to rewrite a good part of the code base to use libdrizzle as well as the new storage engine interface, so they will be somewhat separate projects -- though I'll do my best to keep them the same single project.
I want to thank Monty Widenius and Antony Curtis for fixing some small bugs in order to get this merged!
If anyone has a feature request, just let me know.
The post on this news can also be found at
http://askmonty.org/wiki/index.php/MariaDB_versus_MySQL
November 2 2009, 16:25:21 UTC 2 years ago
Another bug fix...
I have a fix for another small bug, I will push/publish it soon.November 4 2009, 18:57:36 UTC 2 years ago
Re: Another bug fix...
bug fix pending merge review.Anonymous
November 3 2009, 20:26:59 UTC 2 years ago
Documenation?
Can we hope to see some content here (or wherever) http://askmonty.org/wiki/index.php/ManuaNovember 4 2009, 18:59:33 UTC 2 years ago
Re: Documenation?
The biggest difference you would see (as a user) is that FederatedX will make fewer connections to the remote server and it understands transactions - you can commit/rollback and it would propagate those to the remote server. However, it currently does not do 2-phase-commits... yet.Regards,
Antony
November 5 2009, 12:58:38 UTC 2 years ago
Re: Documenation?
I'd have to talk to the maria team about that. Perhaps just a link to our own federatedX page.The differences with FederatedX are as Antony describes, as well as having a completely different code layout. The connection is abstracted into its own class so as to allow creating other adaptors (drizzle, odbc, etc). Also, this project will move forward as opposed to the current one which doesn't. There are several bugs in the old federated that are fixed in FederatedX.