[KUDU-1759] Tablet/table unable to serve write after node crash and restart.

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Cannot Reproduce
• Affects Version/s: 1.0.0, 1.0.1
• Fix Version/s: NA
• Assignee: Unassigned
• Reporter: zhangsong
• Created: 24/Nov/16 09:30
• Updated: 02/Jun/20 13:47
• Resolved: 02/Jun/20 13:47

Description

background:
There are several physical node crash in cluster of jd.com, after restart the kudu-tserver, there are "delete task" task on kudu-master's UI and last minutes (seems will last forever lol~) , mean while the table is unwritable.

Those delete task on master UI is like this:
"Delete Tablet Running 59.2 min b4a24cd8a8b34199aa738607b1c713a9 Delete Tablet RPC for TS=ac6633c779db4cc88ebec15f5e31fb6b"

After use raft-config change tool to "REMOVE_SERVER" these "delete task" disappear.

Issue Links

• is related to: KUDU-1483 in some cases, followers cannot promote to leader.