Global sort documents update (pingcap#16605)

sql-statements/sql-statement-import-into.md

@@ -29,12 +29,12 @@ The `IMPORT INTO` statement lets you import data to TiDB via the [Physical Impor
 
 ### `IMPORT INTO ... FROM FILE` restrictions
 
-- For TiDB Self-Hosted, `IMPORT INTO` supports importing data within 10 TiB. For [TiDB Dedicated](https://docs.pingcap.com/tidbcloud/select-cluster-tier#tidb-dedicated), `IMPORT INTO` supports importing data within 50 GiB.
+- For TiDB Self-Hosted, each `IMPORT INTO` task supports importing data within 10 TiB. If you enable the [Global Sort](/tidb-global-sort.md) feature, each `IMPORT INTO` task supports importing data within 40 TiB.
+- For [TiDB Dedicated](https://docs.pingcap.com/tidbcloud/select-cluster-tier#tidb-dedicated), if your data to be imported exceeds 500 GiB, it is recommended to use TiDB nodes with at least 16 cores and enable the [Global Sort](/tidb-global-sort.md) feature, then each `IMPORT INTO` task supports importing data within 40 TiB. If your data to be imported is within 500 GiB or if the cores of your TiDB nodes are less than 16, it is not recommended to enable the [Global Sort](/tidb-global-sort.md) feature.
 - The execution of `IMPORT INTO ... FROM FILE` blocks the current connection until the import is completed. To execute the statement asynchronously, you can add the `DETACHED` option.
 - Up to 16 `IMPORT INTO` tasks can run simultaneously on each cluster (see [TiDB Distributed eXecution Framework (DXF) usage limitations](/tidb-distributed-execution-framework.md#limitation)). When a cluster lacks sufficient resources or reaches the maximum number of tasks, newly submitted import tasks are queued for execution.
 - When the [Global Sort](/tidb-global-sort.md) feature is used for data import, the value of the `THREAD` option must be at least `16`.
 - When the [Global Sort](/tidb-global-sort.md) feature is used for data import, the data size of a single row after encoding must not exceed 32 MiB.
-- When the Global Sort feature is used for data import, if the target TiDB cluster is deleted before the import task is completed, temporary data used for global sorting might remain on Amazon S3. In this case, you need to delete the residual data manually to avoid increasing S3 storage costs.
 - All `IMPORT INTO` tasks that are created when [TiDB Distributed eXecution Framework (DXF)](/tidb-distributed-execution-framework.md) is not enabled run directly on the nodes where the tasks are submitted, and these tasks will not be scheduled for execution on other TiDB nodes even after DXF is enabled later. After DXF is enabled, only newly created `IMPORT INTO` tasks that import data from S3 or GCS are automatically scheduled or failed over to other TiDB nodes for execution.
 
 ### `IMPORT INTO ... FROM SELECT` restrictions
@@ -184,10 +184,6 @@ For TiDB Self-Hosted, `IMPORT INTO ... FROM FILE` supports importing data from f
 
 ### Global Sort
 
-> **Warning:**
->
-> The Global Sort feature is experimental. It is not recommended to use it in production environments.
-
 `IMPORT INTO ... FROM FILE` splits the data import job of a source data file into multiple sub-jobs, each sub-job independently encoding and sorting data before importing. If the encoded KV ranges of these sub-jobs have significant overlap (to learn how TiDB encodes data to KV, see [TiDB computing](/tidb-computing.md)), TiKV needs to keep compaction during import, leading to a decrease in import performance and stability.
 
 In the following scenarios, there can be significant overlap in KV ranges:
@@ -196,7 +192,7 @@ In the following scenarios, there can be significant overlap in KV ranges:
     - `IMPORT INTO` splits sub-jobs based on the traversal order of data files, usually sorted by file name in lexicographic order.
 - If the target table has many indexes, or the index column values are scattered in the data file, the index KV generated by the encoding of each sub-job will also overlap.
 
-When the [TiDB Distributed eXecution Framework (DXF)](/tidb-distributed-execution-framework.md) is enabled, you can enable [Global Sort](/tidb-global-sort.md) by specifying the `CLOUD_STORAGE_URI` option in the `IMPORT INTO` statement or by specifying the target storage address for encoded KV data using the system variable [`tidb_cloud_storage_uri`](/system-variables.md#tidb_cloud_storage_uri-new-in-v740). Note that currently, only S3 is supported as the Global Sort storage address. When Global Sort is enabled, `IMPORT INTO` writes encoded KV data to the cloud storage, performs Global Sort in the cloud storage, and then parallelly imports the globally sorted index and table data into TiKV. This prevents problems caused by KV overlap and enhances import stability.
+When the [TiDB Distributed eXecution Framework (DXF)](/tidb-distributed-execution-framework.md) is enabled, you can enable [Global Sort](/tidb-global-sort.md) by specifying the `CLOUD_STORAGE_URI` option in the `IMPORT INTO` statement or by specifying the target storage address for encoded KV data using the system variable [`tidb_cloud_storage_uri`](/system-variables.md#tidb_cloud_storage_uri-new-in-v740). Currently, Global Sort supports using Amazon S3 as the storage address. When Global Sort is enabled, `IMPORT INTO` writes encoded KV data to the cloud storage, performs Global Sort in the cloud storage, and then parallelly imports the globally sorted index and table data into TiKV. This prevents problems caused by KV overlap and enhances import stability and performance.
 
 Global Sort consumes a significant amount of memory resources. Before the data import, it is recommended to configure the [`tidb_server_memory_limit_gc_trigger`](/system-variables.md#tidb_server_memory_limit_gc_trigger-new-in-v640) and [`tidb_server_memory_limit`](/system-variables.md#tidb_server_memory_limit-new-in-v640) variables, which avoids golang GC being frequently triggered and thus affecting the import efficiency.
 

system-variables.md

@@ -1625,9 +1625,9 @@ mysql> SELECT job_info FROM mysql.analyze_jobs ORDER BY end_time DESC LIMIT 1;
 
 ### tidb_cloud_storage_uri <span class="version-mark">New in v7.4.0</span>
 
-> **Warning:**
+> **Note:**
 >
-> This feature is experimental. It is not recommended that you use it in the production environment. This feature might be changed or removed without prior notice. If you find a bug, you can report an [issue](https://github.com/pingcap/tidb/issues) on GitHub.
+> Currently, the [Global Sort](/tidb-global-sort.md) process consumes a large amount of computing and memory resources of TiDB nodes. In scenarios such as adding indexes online while user business applications are running, it is recommended to add new TiDB nodes to the cluster and set the [`tidb_service_scope`](/system-variables.md#tidb_service_scope-new-in-v740) variable of these nodes to `"background"`. In this way, the distributed framework schedules tasks to these nodes, isolating the workload from other TiDB nodes to reduce the impact of executing backend tasks such as `ADD INDEX` and `IMPORT INTO` on user business applications.
 
 - Scope: GLOBAL
 - Persists to cluster: Yes

tidb-global-sort.md

@@ -8,9 +8,10 @@ summary: Learn the use cases, limitations, usage, and implementation principles
 
 # TiDB Global Sort
 
-> **Warning:**
+> **Note:**
 >
-> This feature is experimental. It is not recommended that you use it in the production environment. This feature might be changed or removed without prior notice. If you find a bug, you can report an [issue](https://github.com/pingcap/tidb/issues) on GitHub.
+> - Currently, the Global Sort process consumes a large amount of computing and memory resources of TiDB nodes. In scenarios such as adding indexes online while user business applications are running, it is recommended to add new TiDB nodes to the cluster and set the [`tidb_service_scope`](/system-variables.md#tidb_service_scope-new-in-v740) variable of these nodes to `"background"`. In this way, the distributed framework schedules tasks to these nodes, isolating the workload from other TiDB nodes to reduce the impact of executing backend tasks such as `ADD INDEX` and `IMPORT INTO` on user business applications.
+> - When the Global Sort feature is used, it is recommended to use TiDB nodes with at least 16 cores of CPU and 32 GiB of memory to avoid OOM.
 
 > **Note:**
 >
@@ -20,7 +21,7 @@ summary: Learn the use cases, limitations, usage, and implementation principles
 
 The TiDB Global Sort feature enhances the stability and efficiency of data import and DDL (Data Definition Language) operations. It serves as a general operator in the [TiDB Distributed eXecution Framework (DXF)](/tidb-distributed-execution-framework.md), providing a global sort service on cloud.
 
-The Global Sort feature currently only supports using Amazon S3 as cloud storage. In future releases, it will be extended to support multiple shared storage interfaces, such as POSIX, enabling seamless integration with different storage systems. This flexibility enables efficient and adaptable data sorting for various use cases.
+Currently, the Global Sort feature supports using Amazon S3 as cloud storage.
 
 ## Use cases