[SAMZA-2423] Heartbeat failure leads the to container shutdown with exit code 0

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Assignee: Abhishek Shivanna
• Reporter: Abhishek Shivanna
• Created: 20/Dec/19 02:46
• Updated: 21/Dec/19 02:24
• Resolved: 21/Dec/19 02:24

Description

Background: A Samza container periodically calls the AM through a HeartbeatMonitor while the container is running. If a container can't contact the AM because of a network issue or the AM moved hosts, the container will shutdown afterwards and is expected to restart afterwards.

The issue is that the current code (ContainerLaunchUntil.java) returns with a status exit code of 0 leading the job coordinator to not reschedule the container.

 

Issue Links

• links to: GitHub Pull Request #1240