Where the information about the details of MediaWiki web service API? Quoting Fernando Correia <fernandoacorreia(a)gmail.com>om>: ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.

I am didn't heard about the Web service before this thread... Quoting Rob Church <robchur(a)gmail.com>om>: ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Ittay Dror wrote: Can you explain for the gallery what's non-RESTy about the current draft api? - -- brion vibber (brion @ pobox.com / brion @ wikimedia.org)

Russ Lavoie wrote: documentation. How do you synchronize the MySQL? Or are you using a single remote MySQL server that all servers use. -Jim _______________________________________________ MediaWiki-l mailing list MediaWiki-l(a)lists.wikimedia.org http://lists.wikimedia.org/mailman/listinfo/mediawiki-l

Hello Brion, Can you explain why it will not work? I know the is counter issue that can be tweak. I have good experience on internal LAMP application and tough abut give it a try on mediawiki. Addady Brion Vibber wrote:

So what I am getting from all of these posts is to use rsync for the frontend and mysql clustering for the backend... Is there any other way to get this working with a better solution other than rsync? Thanks Russ -----Original Message----- From: mediawiki-l-bounces(a)lists.wikimedia.org [mailto:mediawiki-l-bounces@lists.wikimedia.org] On Behalf Of Rob Church Sent: Wednesday, January 24, 2007 12:06 PM To: MediaWiki announcements and site admin list Subject: Re: [Mediawiki-l] Load Balancing? On 24/01/07, Sullivan, James (NIH/CIT) [C] <sullivan(a)mail.nih.gov> wrote: Wikimedia uses a single master and multiple slaves in a replication environment. Our biggest wiki, the English Wikipedia, has a separate master to that of the other wikis, and a separate slave cluster, I believe. Rob Church _______________________________________________ MediaWiki-l mailing list MediaWiki-l(a)lists.wikimedia.org http://lists.wikimedia.org/mailman/listinfo/mediawiki-l

Let me explain a little more, rsync has to tally up all the files in the directories you will be sharing between the two servers which puts an unnecessary load on the system (plus it isn't always up-to-date on both servers). I would like the updates to be real time. Any suggestions? Thanks, Russ -----Original Message----- From: mediawiki-l-bounces(a)lists.wikimedia.org [mailto:mediawiki-l-bounces@lists.wikimedia.org] On Behalf Of Brion Vibber Sent: Wednesday, January 24, 2007 1:16 PM To: MediaWiki announcements and site admin list Subject: Re: [Mediawiki-l] Load Balancing? -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Russ Lavoie wrote: way It might help if we knew what "better" meant. What requirements do you have for deploying updated software that are not met by something like rsync? - -- brion vibber (brion @ pobox.com / brion @ wikimedia.org) _______________________________________________ MediaWiki-l mailing list MediaWiki-l(a)lists.wikimedia.org http://lists.wikimedia.org/mailman/listinfo/mediawiki-l

On 1/24/07, Brion Vibber <brion(a)pobox.com> wrote: Just to clarify in case it's unclear to Russ: The mediawiki CONTENT (article pages, etc) is not hosted on the webservers. It's in the database server(s). Except for the images, and handling that via a NFS server works fine (or, worked fine for me, with my testing, I don't run it in production). The webservers only need to be synchronized if you change something in the MediaWiki configuration. Rsync or the equivalent will work fine (or CacheFS, or...). -- -george william herbert george.herbert(a)gmail.com

Hi! Indeed, there's no better example of scaled out mediawiki than Wikipedia. Even for single wiki, say English Wikipedia we use following distribution: - Multiple tiers of Squids - Multiple Apache/PHP servers, running MediaWiki - Multiple memcached servers for object cache and various not-very- persistent storage - Multiple Lucene hosts serving search - Multiple clusters (groups) of external storage nodes, each consisting of small master-slaves replication system carrying a subset of texts - Master database - Slave databases for general DB use ( can be promoted to master, if needed) - Slave databases allocated for specific DB use (as Ariel serves watchlists) - Slave databases extended/built for specific DB use (extended indexing on db6 to allow faster contributions access) - Multiple hosts running job queue (delayed tasks) - Multiple load balancers between various components. Once you need to scale more than English wikipedia, we can discuss about additional possibilities to split some of the tasks. All the code for above load balancing is inside mediawiki. Cheers, -- Domas Mituzas -- http://dammit.lt/ -- [[user:midom]]

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Tim Starling wrote: Two additional notes: 1) Be sure to keep the servers' clocks in sync! A standard NTP system will do. 2) PHP's session storage is used to handle login sessions. Session data must be stored in a common location unless your load balancing system is "sticky" (sending a given client always to the same server). A networked filesystem will usually do fine, or we have the 'sessions in memcached' sample implementation of storing session data in a shared memcached cloud. - -- brion vibber (brion @ pobox.com / brion @ wikimedia.org)