[BUILD-UNSTABLE]: Job 'IoTDB/IoTDB-pip-new/master [master] [422]'

[Apache Jenkins Server]

BUILD-UNSTABLE: Job 'IoTDB/IoTDB-pip-new/master [master] [422]':

Check console output at "https://ci-builds.apache.org/job/IoTDB/job/IoTDB-pip-new/job/master/422/";>IoTDB/IoTDB-pip-new/master
 [master] [422]"

[jira] [Created] (IOTDB-6266) Add the ability to flush syncIndex and update reader periodically for IoTConsensus

[Xinyu Tan (Jira)]

Xinyu Tan created IOTDB-6266:


 Summary: Add the ability to flush syncIndex and update reader 
periodically for IoTConsensus
 Key: IOTDB-6266
 URL: https://issues.apache.org/jira/browse/IOTDB-6266
 Project: Apache IoTDB
  Issue Type: Improvement
Reporter: Xinyu Tan
Assignee: Xinyu Tan
 Attachments: image-2023-12-11-17-43-20-151.png

After the PR is merged , The safeDeletedSearchIndex passed by IoTConsensus to 
the wal reader is no longer the syncIndex that has been synchronized, but the 
SyncIndex that has been flushed to disk.

When a leader migration is triggered, the problem may occur that the wal of the 
old leader can never be deleted, resulting in a pile-up of the wal.

!image-2023-12-11-17-43-20-151.png!

To solve this problem, iotconsensus can add a way to flush syncIndex and update 
reader periodically to avoid the accumulation of the old leader's log after the 
leader switch



--
This message was sent by Atlassian Jira
(v8.20.10#820010)