[CASSANDRA-15040] Join / Decommission Cascadra nodes streaming failed

• Type: Improvement
• Status: Open
• Priority: Normal
• Resolution: Unresolved
• Fix Version/s: 3.11.x
• Component/s: Cluster/Gossip, Consistency/Bootstrap and Decommission

Environment

Ubuntu 16.04.3 LTS

AWS EC2 

• Assignee: Unassigned
• Reporter: SungJae Yun
• Created: 04/Mar/19 03:13
• Updated: 08/Jul/19 16:56

Description

I am having some problem with Cassandra 3.10 that when I try to join/decommission node, it cannot join/decommission the cluster with the "stream failed" exception, then it stops running. Util that time, node stays in JOINING state.
 
It succeeds in the following cases.
 
case 1.

drop index > node join/decommission SUCCEED > create index > node join/decommission FAIL

case 2.

table data backup (copy to command) > truncate table > data recovery (copy from command) > node join/decommission SUCCEED > New DATA insert from service > node join/decommission FAIL

case 3.

table data backup (copy to command) > drop table > create table >data recovery (copy from command) >node join/decommission SUCCEED > New DATA insert from service > node join/decommission FAIL
 
Unfortunately, I do not know about new insert queries from the service.
 
Q1. Can I check queries from the service session? 
 
Q2. Do you have any idea why stream fails in cassandra?