Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.
Original topic: Tidb-lightning导入报错fetch table columns failed with no retry
【TiDB Usage Environment】Production
【TiDB Version】v6.1
【Encountered Problem】
【Reproduction Path】
Currently executing data import to TiDB 6.1
【Problem Phenomenon and Impact】
- Error during import process: Error 1105: message: “raft entry is too large, region 90791752, entry size 8960540” raft_entry_too_large
- Based on community posts, attempted to modify raftstore.raft-entry-max-size configuration to 128M and reloaded the TiKV nodes;
- Re-ran lightning, encountered error “fetch table columns failed: sql: Scan error on column index 3, name "NEXT_GLOBAL_ROW_ID": converting driver.Value type uint8 ("12844501496393138768") to a int64: value out of range”], this error occurs at the start of the re-run and cannot be skipped.
How should I handle this, teacher?
During the reload of TiKV, there was an error indicating that the TiKV node took more than 2 minutes to start. Checking the cluster status showed “Disconnected,” but after a few seconds, the status returned to normal.
tidb:
performance.txn-entry-size-limit: 12582912
tikv:
raftstore.raft-entry-max-size: '16MB'
Try adjusting like this and reload the TiDB and TiKV nodes.
After adjusting and reloading TiDB and TiKV, the error still occurs:
[ERROR] [main.go:103] [“tidb lightning encountered error stack info”] [error=“fetch table columns failed: sql: Scan error on column index 3, name "NEXT_GLOBAL_ROW_ID": converting driver.Value type uint8 ("12844501496393138768") to a int64: value out of range”]
From the error message, it appears that the field value “12844501496393138768” for NEXT_GLOBAL_ROW_ID exceeds the maximum value for int64. Try changing the field type to an unsigned bigint.
From the error message, it is not clear which table’s field it is, and it is not a business table’s field either.
Which version of Lightning are you using?
|
username: TiDBer_spike
|
username: