[CASSANDRA-16997] Handle timestamp wraparound for internode messages

• Type: Bug
• Status: Resolved
• Priority: Normal
• Resolution: Fixed
• Fix Version/s: 4.0.2, 4.1-alpha1, 4.1
• Component/s: Messaging/Internode
• Bug Category: Availability - Response Crash
• Severity: Normal
• Complexity: Normal
• Discovered By: Adhoc Test
• Platform: All
• Since Version: 4.0-alpha1

Source Control Link

https://github.com/apache/cassandra/commit/84ec1dc97d6358bd569d5467cb150abd0fc8939b

Test and Documentation Plan

new test + cci run

• Assignee: Marcus Eriksson
• Reporter: Marcus Eriksson

Authors

Marcus Eriksson

Reviewers

Jon Meredith

• Created: 27/Sep/21 08:29
• Updated: 27/May/22 19:25
• Resolved: 30/Sep/21 09:29

Description

In internode messaging we send the lower 4 bytes of the timestamp and then correct on receiving side if the local time has wrapped (every 50 days), but we currently don't correct the time if the sending side wrapped before us.

This can cause us to use the wrong timestamp for the message and if the local System.nanoTime is < 50 days, we get a negative creation time which causes an AssertionError in MonitorableImpl