Really good point about the firewall loophole, thanks for bringing it up.
The code that I wrote is very much the bridge daemon you suggested. So I
guess it just needs to remain living separately from Thrift.
-Daniel
On 9/10/10 12:05 PM, Time Less wrote:
If it were to be included in HBase in any fashion, I think the best way
that could be done would be to modify the Thrift server to have an optional
HTTP port it could listen on for health checks. That way it would be easy
for a non-thrift client such as Nagios or a load balancer to still query the
Thrift server for general health.
Don't do that. If for some reason the main connection port can't be
contacted, but the "health check" port can, then Nagios doesn't alert but
service is down. Example: someone firewalls off Thrift port but not HTTP
healthcheck port.
The server daemon should not be able to tell what's connecting to it is an
actual client trying to do real work or a monitoring service (other than
that monitoring services tend to ask Really Easy Questions).
Easiest way to accomplish this is create a bridge daemon between the
protocol the Nagios/loadBalancer speaks and the protocol the server daemon
speaks. In this case, HTTP-->Thrift.
