I’ve learned that a recommended :onyx.peer/coordinator-barrier-period-ms setting for production is not 1 second but somewhere along 30 to 60 seconds. Checkpointing will happen too often and can fill up pretty quickly

Hi. Has anyone successfully used :onyx.messaging.aeron/media-driver-dir setting?

We’ve been trying to track down some exceptions we’re getting from aeron in our onyx 0.14 cluster (we don’t get them in our 0.9 cluster). Specifically it looks like we’re getting ConductorServiceTimeoutExceptions though there are other errors we get around the same time, for example Aeron Status Channel Errors

One thing I noticed this morning was that there are hundreds of client-conductor threads on the replica I dug in to, there are only 11 vpeers on that replica, is that normal? Our 0.9 cluster replicas have far less, 11 threads for 20 vpeers