[CASSANDRA-4484] Drain causes incorrect error messages: "Stream took more than 24H to complete; skipping"

• Type: Bug
• Status: Resolved
• Priority: Low
• Resolution: Fixed
• Fix Version/s: 1.1.4
• Severity: Low
• Assignee: Christopher Porter
• Reporter: Christopher Porter

Authors

Christopher Porter

Reviewers

Jonathan Ellis

• Created: 02/Aug/12 16:41
• Updated: 16/Apr/19 09:32
• Resolved: 02/Aug/12 17:57

Description

After calling drain on a node, there are a bunch of incorrect error messages in the cassandra log file: "Stream took more than 24H to complete; skipping".

The problem is in MessagingService.waitForStreaming. It is logging an error if ThreadPoolExecutor.awaitTermination returns true, but if a timeout happens it returns false. See http://docs.oracle.com/javase/7/docs/api/java/util/concurrent/ThreadPoolExecutor.html#awaitTermination%28long,%20java.util.concurrent.TimeUnit%29