Hi All, 1) Polling transports like VFS gets into trouble when we run them in a cluster, as they might end of processing many copies of the same message
2) Our answer now is use Pinned server attribute in the proxy services 3) However, now HA is a problem as if the pinned server died, that proxy is lost. Just like the Zookeeper based task implementation, I think we should support Zookeeper/ Hazecast based support for a proxy service to say that there should be N instances of that proxy service running. WDYT? --Srinath -- ============================ Srinath Perera, Ph.D. Director, Research, WSO2 Inc. Visiting Faculty, University of Moratuwa Member, Apache Software Foundation Research Scientist, Lanka Software Foundation Blog: http://srinathsview.blogspot.com/ Photos: http://www.flickr.com/photos/hemapani/ Phone: 0772360902
_______________________________________________ Architecture mailing list [email protected] https://mail.wso2.org/cgi-bin/mailman/listinfo/architecture
