Sudharma/Tammo, 1) How do we plan to decide which is the master node in the cluster?
2) Don't we need to stall the Deployment Pollers in the slave nodes? Suggestion: I am not sure whether do we need Master-SLaves. Why not give every node in the cluster the same status (Active-Active). When a new deployment is made, the load balancer can push it to any of the available nodes. That node will probably acquire a distributed lock on the deployment unit and acts as master for that deployment. This ensures optimum usage of the cluster nodes. Probably no static configuration of Master-Slave in the load balancer nor in the hazelcast. On Tue, May 19, 2015 at 4:31 PM, Tammo van Lessen <tvanles...@gmail.com> wrote: > Hi Sudharma, > > Electing one master that is able to deploy is the right approach. Since the > metadata of a deployment is written to the database and the database is > shared among all nodes, sending notifications is not necessary. This is all > under the assumption that all nodes share the same deployment folder, e.g. > using SMB or NFS. > > HTH, > Tammo > > On Thu, May 14, 2015 at 11:59 AM, sudharma subasinghe < > suba...@cse.mrt.ac.lk > > wrote: > > > Hi, > > > > In the cluster, every node shares the database and deployment directory. > To > > avoid race condition I am going to introduce master and slave > > configuration. Here master node will deploy the process and write the > > version number to the database. Until master node finishes its process > all > > the salve nodes will wait. After finishing the master node's task it will > > send a custom message to slave nodes through the cluster. Then they can > get > > the version number from database and will read the deployed process to > > deploy it in salve nodes. Please give me ideas on this. > > > > Thanks, > > Sudharma > > > > > > -- > Tammo van Lessen - http://www.taval.de >
