jackrabbit-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Lukas Kahwe Smith <...@pooteeweet.org>
Subject Re: JR - Load Balancing & Fault Tolerance
Date Fri, 16 Sep 2011 12:56:43 GMT

On Sep 16, 2011, at 14:50 , Francisco Carriedo Scher wrote:

> Hi there,
> 
> i have been checking the Wiki section about clustering and users' mailing
> listthe  as well. I am missing some info about what clustering Jackrabbit
> will provide and what not.
> 
> It will provide _fault tolerance_ through journaling but, is this feature
> really mature? I read that there are issues regarding to de-synchronisation
> between journaling and file datastore state, too frequent sync delay causes
> issues too, it is needed to fine tune the configuration enlarging cache
> size...

at liip we are using rdbms data store (with mysql) and have come up with the following solution
to be able to globally distribute our cluster nodes:
http://blog.liip.ch/archive/2011/05/04/how-to-make-jackrabbit-globally-distributable-fail-safe-and-scalable-in-one-go.html

> It will _not_ provide built in load balancing so, the only option i see so
> far is to access Jackrabbit through Webdav and use Apache webserver to
> balance the requests along all the clustered nodes. Again, is this the best
> way and are there any critical issues that i must consider?

i should note that we access jackrabbit via davex and so we use a standard load balancer to
direct the davex requests. our client lib adds a random header to each request which the load
balancer uses to determine which jackrabbit node to redirect the traffic to.

regards,
Lukas Kahwe Smith
mls@pooteeweet.org




Mime
View raw message