Hi, I am working on a hadoop based log archival solution where logs are generated by telecom Routers. I have very limited knowledge of the routers but I know that routers will send syslogs to given IP address and port. If we want to use Flume for collecting these syslog, we will run flume agent having Syslog UDP as source and running on the same server where router will send logs. Assuming router can send log messages to only single IP, how can we achieve fail-over for flume agent in such cases ?
This is something urgent and any help would be greatly appreciated. Regards Abhijeet
