provide service �→ normally. backend = "local" ### Set the temporary storage directory for the sorted Key-Value files. The �→ directory must be empty, and the storage space must be greater than �→ the different from `data-source- �→ dir` and use flash storage, which can use I/O exclusively. sorted-kv-dir = "/mnt/ssd/sorted-kv-dir" [mydumper] ### The path that stores the snapshot file. data-source-dir = "${s3_path}" exported). SSD is recommended. • During the import, TiDB Lightning needs temporary space to store the sorted key- value pairs. The disk space should be enough to hold the largest single table from the data

updates. However, the disadvantage is also obvious. As the primary key or unique index needs to be sorted, a larger offset consumes more computing resources, especially in the case of a large volume of data factors are as follows: • Source files • Whether the data within a single file is sorted by the primary key. Sorted data can achieve optimal import performance. • Whether overlapping primary keys or non-null unique indexes exist between multiple source files. If the generated files are globally sorted, they can be distributed into different TiDB Lightning instances based on ranges to achieve optimal

updates. However, the disadvantage is also obvious. As the primary key or unique index needs to be sorted, a larger offset consumes more computing resources, especially in the case of a large volume of data factors are as follows: • Source files • Whether the data within a single file is sorted by the primary key. Sorted data can achieve optimal import performance. • Whether overlapping primary keys or non-null unique indexes exist between multiple source files. If the generated files are globally sorted, they can be distributed into different TiDB Lightning instances based on ranges to achieve optimal

· · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · 6729 17.15.2 Static Sorted Table / Sorted String Table (SST) · · · · · · · · · · · · · · · · · · · 6730 17.15.3 Store · · · · · · · updates. However, the disadvantage is also obvious. As the primary key or unique index needs to be sorted, a larger offset consumes more computing resources, especially in the case of a large volume of data +----+----------+---------------------+ 3 rows in set (0.15 sec) The three terms in the search results are sorted by their respective distance from the queried vector: the smaller the distance, the more relevant

return value contains pagination information such as offset, total pages, and whether the results are sorted. 176 { "content": [ { "coins": 200, "goods": 10, "id": 1 }, { "coins": 0, "goods": 30, "id": "pageSize": 2, "paged": true, "sort": { "empty": true, "sorted": false, "unsorted": true }, "unpaged": false }, "size": 2, "sort": { "empty": true, "sorted": false, "unsorted": true }, "totalElements": 4 "goods":10},{"id":2,"coins":0,"goods":30}]," �→ pageable":{"sort":{"empty":true,"unsorted":true,"sorted":false}," �→ offset":0,"pageNumber":0,"pageSize":2,"paged":true,"unpaged":false}," �→ last":false

updates. However, the disadvantage is also obvious. As the primary key or unique index needs to be sorted, a larger offset consumes more computing resources, especially in the case of a large volume of data +----+----------+---------------------+ 3 rows in set (0.15 sec) The three terms in the search results are sorted by their respective distance from the queried vector: the smaller the distance, the more relevant 6469335836410557 - text: "tree", distance: 0.798545178640937 The three terms in the search results are sorted by their respective distance from the queried vector: the smaller the distance, the more relevant

is empty and that the �→ disk capacity is greater than the size of the dataset to be �→ imported. sorted-kv-dir = "/path/to/local-temp-dir" 4. Configure the file routing. [mydumper] no-schema = true S3: ./tidb-lightning --tidb-port=4000 --pd-urls=127.0.0.1:2379 --backend= �→ local --sorted-kv-dir=/tmp/sorted-kvs \ -d 's3://my-bucket/sql-backup?region=us-west-2' • Use TiDB Lightning to import data mode): ./tidb-lightning --tidb-port=4000 --pd-urls=127.0.0.1:2379 --backend= �→ local --sorted-kv-dir=/tmp/sorted-kvs \ -d 's3://my-bucket/sql-backup?force-path-style=true&endpoint=http �→ ://10.154.10

�→ processing speed. When memory resources are insufficient, parallel �→ HashAgg spills temporary sorted data to disk, avoiding potential OOM �→ risks caused by excessive memory usage. This improves query updates. However, the disadvantage is also obvious. As the primary key or unique index needs to be sorted, a larger offset consumes more computing resources, especially in the case of a large volume of data factors are as follows: • Source files • Whether the data within a single file is sorted by the primary key. Sorted data can achieve optimal import performance. • Whether overlapping primary keys or

updates. However, the disadvantage is also obvious. As the primary key or unique index needs to be sorted, a larger offset consumes more computing resources, especially in the case of a large volume of data factors are as follows: • Source files • Whether the data within a single file is sorted by the primary key. Sorted data can achieve optimal import performance. • Whether overlapping primary keys or non-null unique indexes exist between multiple source files. If the generated files are globally sorted, they can be distributed 715 into different TiDB Lightning instances based on ranges to achieve

updates. However, the disadvantage is also obvious. As the primary key or unique index needs to be sorted, a larger offset consumes more computing resources, especially in the case of a large volume of data factors are as follows: • Source files • Whether the data within a single file is sorted by the primary key. Sorted data can achieve optimal import performance. • Whether overlapping primary keys or non-null unique indexes exist between multiple source files. If the generated files are globally sorted, they can be distributed 710 into different TiDB Lightning instances based on ranges to achieve