A distributed system that performs a likelihood scan of event directions for IceCube real-time alerts using CPU cluster(s) and queue-based message passing.
Wait only on the server. We don't actually care about the clients, if they fail or not. Now, if all the clients fail, then the sever times out, the test fails, and we can look at client logs.
This has the effect of allowing one failed client, which forgives some troublesome issues seen occasionally.
Wait only on the server. We don't actually care about the clients, if they fail or not. Now, if all the clients fail, then the sever times out, the test fails, and we can look at client logs.
This has the effect of allowing one failed client, which forgives some troublesome issues seen occasionally.