[IOTDB-1183] Optimize Query Performance in Cluster

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Duplicate
• Fix Version/s: 0.12.0
• Component/s: Core/Cluster
• Assignee: Yifu Zhou
• Reporter: Yifu Zhou
• Created: 04/Mar/21 09:25
• Updated: 25/Aug/21 08:17
• Resolved: 10/Mar/21 02:30

Description

In IoTDB cluster version, query performance is not good enough, especially if we need to read the data from remote nodes. We need to reduce unnecessary operations and improve the query performance in cluster version ASAP.

1. ManagedSeriesReader will sync leader for each timeseries. For example, if query select s_1,s_2 from root.to.path where time < 5, it will do syncLeaderWithConsistencyCheck twice. Actually it only needs to do check once for each query. I will fix it.