Hi Rich, thanks for response, some details about "route manipulation through user interaction": we splitted our routes in a kind of templates for protocol and transformation. So user can create new routes or changes route parameters such as uri, transformation or timer settings through web frontend an start them. Routeconfiguration will be persisted in our application database.
New or changed routes should be available on all nodes. Most of our routes did data integration robs. Regards Robin -----Ursprüngliche Nachricht----- Von: Rich Newcomb [mailto:rich.newc...@gmail.com] Gesendet: Donnerstag, 27. Juni 2013 16:40 An: users@camel.apache.org Betreff: Re: [Best practise question] Running Camel in JBoss cluster Hi Robin, You should be in good shape with clustered quartz scheduling and the servlet endpoints fronted by a load balancer. Could you provide some more information about the "route manipulation through user interaction"? Describing this requirement more fully may lead to more and better responses. On Thu, Jun 27, 2013 at 9:20 AM, Lutter, Robin <robin.lut...@t-systems.com> wrote: > Hi @all, > > we plan to run camel in clustered environment. Our target platform is > JBoss 7.1.3 (or Wildfly EAP 6.0) with following challenges: > > 1. most of our routes (httpConsumer, fileConsumer, ftpConsumer) are > triggered using quartz: a triggered route should run on one node at > time 2. some servlet endpoints, which are load balanced 3. route > manipulation through user interaction during runtime on one node must > be distributed to all other nodes 4. multiple nodes for failover and > work balancing > > Our current approach is: > Running quartz clustered should be no problem, isnt' it? > Every node starts it's own camel context should with working routes. > So there all servlet endpoints are accessible on all nodes. > > My question is there a good source or real life experience for running > camel in distributed environment. > How to sync different camel context? > > Thanks Robin
