Hello,
I was assigned a POC where a cluster should not have a single point of
failure -- more specifically, each node should be assigned a different role.
This means that loading the cache, querying, storage, etc. would be seen as
their own individual roles and therefore kept separate, hosting on different
ports. 
For example, a node would be assigned the role of querying the cache;
ideally, a port would be set up so that a client would connect to that port. 

I was just wondering if (1) this was possible, and (2) if possible, how
would I go about it? 

Right now, I have  this XML file
<http://apache-ignite-users.70518.x6.nabble.com/file/t3022/example-config.xml>  
where I start a cluster from a remote machine; I have a  one C# project
<http://apache-ignite-users.70518.x6.nabble.com/file/t3022/Program.cs>  
that connects to the cluster, creates a cache, and loads entries from an
underlying RDBMS. Additionally, I have   another C# project
<http://apache-ignite-users.70518.x6.nabble.com/file/t3022/ThinClientSql.cs>  
that connects a thin client to the cluster and queries the cache once
loaded. 
If I were to implement nodes operating separately with their own roles,
would I need to have different XML files with the same [remoteHost] but
different ports under the addresses property for each XML file? (i.e. XML
file #1 has [remoteHost], [remoteHost]:10800; XML file #2 has [remoteHost],
[remoteHost]:47500..47509).




--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/

Reply via email to