Hello. The Hilary cluster setup is a breeze. nicely done.
I'm unsure about config.cassandra in Hilary/config.js. I setup a cassandra node on another server and run it in a cluster with the cassandra node running on the same server as the application server/Hilary. Even with the cassandra cluster and its replication setup done, do I still need to add the remote server (server not running on same server as Hilary) to the config.cassandra.hosts list? Thanks for the help! On 15 November 2013 14:18, Branden Visser <mrvis...@gmail.com> wrote: > Hi Kabelo, > > We don't have formal detailed documentation on this, however there are > bootcamp slides available which give a high-level overview of a > clustered application setup [1]. > > For all components other than Hilary, I would recommend becoming > familiar with their own clustering abilities and reference their > documentation for setup. > > For Hilary, it's important to note the "specialization" of Hilary > nodes into one or more of the following roles: > > 1. Application Request Handler > 2. Activity Processor > 3. Search Indexer > 4. Preview Processor > > For #1, you'll see that Nginx proxies web requests to the application > nodes. You simply need to configure Nginx appropriately to use all > your application request handler nodes into the backend pools in > Nginx. Every Hilary node will always be prepared to handle web > requests after startup. > > For #2-4, they receive their requests through RabbitMQ messages, so as > long as they are configured to talk to RabbitMQ (config.mq in > config.js), they are capable of being distributed and handling these > tasks. Therefore, in order to deploy a specialized node, you simply > need to configure its RabbitMQ connection (*all* Hilary nodes should > have this configured any way), and enable / disable the type of work > you would like that node to handle in config.js: > > 2. Set "config.search.processIndexJobs" to true to indicate the node > should subscribe to search indexing tasks > 3. Set "config.activity.processActivityJobs" to true to indicate the > node should subscribe to activity routing tasks and to aggregate > activity > 4. Set "config.previews.enabled" to true to indicate the node should > subscribe to preview processing tasks > > Note that our puppet scripts for deploying and configuring Hilary > clusters are available [2]. Feel free to look into those to get > specific configuration templates for setting up an OAE cluster. > > Hope that helps, > Branden > > [1] http://www.slideshare.net/nicolaasmatthijs/apereo-oae-bootcamp (Slide > #18) > [2] https://github.com/oaeproject/puppet-hilary > > On Fri, Nov 15, 2013 at 6:19 AM, Kabelo Letsoalo > <kab...@opencollab.co.za> wrote: > > Hi all. where can I find documentation for OAE 2.0.0 cluster setup and/or > > running multiple instances? > > > > ________________________________ > > See OpenCollab email disclaimer at > > http://www.opencollab.co.za/email-disclaimer > > _______________________________________________ > > oae-dev mailing list > > oae-dev@collab.sakaiproject.org > > http://collab.sakaiproject.org/mailman/listinfo/oae-dev > > > -- ------------------------------ See OpenCollab email disclaimer at http://www.opencollab.co.za/email-disclaimer
_______________________________________________ oae-dev mailing list oae-dev@collab.sakaiproject.org http://collab.sakaiproject.org/mailman/listinfo/oae-dev
