[IOTDB-3335] Redundant data recovered by wal in multi-leader consensus

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Component/s: Core/WAL
• Labels: pull-request-available
• Sprint: 2022-5-StorageEngine
• Assignee: Haiming Zhu
• Reporter: Haiming Zhu
• Created: 30/May/22 09:52
• Updated: 31/May/22 13:09
• Resolved: 31/May/22 01:00

Description

Some data will be recovered wrongly because wal is kept in multi-leader consensus, memtable id still starts from 0

SET STORAGE GROUP TO root.sg1;
create aligned timeseries root.sg1.d1(s1 FLOAT encoding=RLE);
insert into root.sg1.d1(time, s1) aligned values(1, 1.0);
restart
insert into root.sg1.d1(time, s1) aligned values(1, 10.0);
insert into root.sg1.d1(time, s1) aligned values(2, 2.0);
restart
insert into root.sg1.d1(time, s1) aligned values(2, 20.0);
insert into root.sg1.d1(time, s1) aligned values(3, 3.0);
restart
insert into root.sg1.d1(time, s1) aligned values(3, 3.0);
insert into root.sg1.d1(time, s1) aligned values(4, 4.0);
restart  

Issue Links

• duplicates: IOTDB-3330 Sequence file has overlapped data after restarting
• links to: GitHub Pull Request #6076