Hi,
Im using Twill to run a batch job on a Yarn cluster. The container gets
killed sometimes because the memory requested was not enough. I see that twill
keeps retrying infinitely. Is there a way I can ask the framework not to
retry ? Looking at the code, I see this
in RunningContainers
if (shouldRetry(exitStatus)) {
LOG.info("Re-request the container {} for exit code {}.", containerId,
exitStatus);
restartRunnables.add(lookup.keySet().iterator().next());
}
Depending on the exit status, we might be restating the containers again and
again.
Sam
