Details
-
Improvement
-
Status: Closed
-
Major
-
Resolution: Fixed
-
None
Description
At present, the executeBatch method in JDBC is executed separately for each statement.
For the cluster version, each statement will be a raft log, which will lead to the cost of the raft log scrambling for locks and the RPC cost between the raft replication groups.
The insertRecords and insertTablets interfaces in session API also have the same problem, for details please see the following links.
The solution is similar to the previous one. The only difference is that the statements in the executeBatch interface may not only insert data, but also have metadata operations(like set storage group, create timeseries and so on). It is necessary to distinguish between metadata operations and data operations.
Attachments
Issue Links
- links to