Hi Smith, I would like to participate in the drizzle project in this year's GSoC. I'm a university student in Shanghai China. I am interested in the storage engine related ideas on the wiki page, especially these two: file system and the column-based archive engine.
The work I'm already done is below: - setup my environment up, both Mac OSX and RedHat Linux. - fix two bugs 541709 and 549914. - get familiar with storage engine interfaces - work out a demo for this file system storage engine. ( lp:~ziminq/drizzle/fs-plugin and blueprint is https://blueprints.launchpad.net/drizzle/+spec/fs-plugin-prototype ) It can work basically. The first example in mysql-filesystem-engine works fine here. But it comes with the following limits: - the specific file has to be placed at your deployment directory, due to the fact drizzle can't pass 'connect string' to the storage engine right now. - it can only read a file. that is to say, only 'select' is supported for this prototype. Though I have already started working on the filesystem plugin idea, I doubt that this plugin is VERY useful to users of drizzle. How do you take this? In this respect, do you think the column-based archive storage engine would benefit drizzle community more than filesystem engine? -ZQ _______________________________________________ Mailing list: https://launchpad.net/~drizzle-discuss Post to : [email protected] Unsubscribe : https://launchpad.net/~drizzle-discuss More help : https://help.launchpad.net/ListHelp
