[IOTDB-604] [Distributed] More robust cluster readers

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Component/s: Core/Cluster
• Labels: cluster pull-request-available query robustness
• Assignee: WangChao
• Reporter: Tian Jiang
• Created: 20/Apr/20 03:27
• Updated: 20/May/20 04:00
• Resolved: 20/May/20 04:00

Description

The current cluster readers only read data from a fixed node so if the targeted node fails, the query is errored and must be retried.

To improve the robustness of queries, the cluster readers should know the data group to query instead of a single node and it should also remember its current position so that after switching to another node, it will not read duplicated data.

Issue Links

• is a child of: IOTDB-68 IoTDB supports distributed version
• links to: GitHub Pull Request #1090
• links to: GitHub Pull Request #1115