On 01/31/2011 09:59 PM, Paul Libbrecht wrote: > > Le 31 janv. 2011 à 19:48, Sergiu Dumitriu a écrit : >>> The interest is that it could allow some of the web requests to be served >>> by a database that is a slave of the main database, in live synchronization >>> with the master MySQL. >> >> You mean that the public wiki reads data from the read-only database, >> which is actually updated by cloning another read-write database updated >> from a private wiki? > > Actually, the goal is the converse: the internal, robot-oriented xwiki would > read the cloned DB. > >> That's an interesting idea for securing a public server. > > could be though you need a filtering mechanism thn. > >> You should also set in xwiki.cfg: >> xwiki.readonly=1 > > why *also* ?
"Set" it so that some of the code inside the platform is aware that the wiki is supposed to be read-only and act accordingly. "Also", because that's not enough to guarantee that no changes will be made to the data, and the database-level read-only is much safer. >> This might reduce some of the errors that you get, although there's a >> lot of code that ignores this setting. It would be nice if you could >> report (on jira) what exactly breaks so that we can fix it properly. > > If you say a lot of code ignores that than I'd better not rely on it I'm > afraid. > Nonetheless, the previous option sounds good. -- Sergiu Dumitriu http://purl.org/net/sergiu/ _______________________________________________ devs mailing list [email protected] http://lists.xwiki.org/mailman/listinfo/devs
