Merge pull request ClickHouse#60550 from ClickHouse/revert-53770-pc

Revert "Userspace page cache"

docs/en/operations/storing-data.md

@@ -275,16 +275,6 @@ Cache profile events:
 
 - `CachedWriteBufferCacheWriteBytes`, `CachedWriteBufferCacheWriteMicroseconds`
 
-## Using in-memory cache (userspace page cache) {#userspace-page-cache}
-
-The File Cache described above stores cached data in local files. Alternatively, object-store-based disks can be configured to use "Userspace Page Cache", which is RAM-only. Userspace page cache is recommended only if file cache can't be used for some reason, e.g. if the machine doesn't have a local disk at all. Note that file cache effectively uses RAM for caching too, since the OS caches contents of local files.
-
-To enable userspace page cache for disks that don't use file cache, use setting `use_page_cache_for_disks_without_file_cache`.
-
-By default, on Linux, the userspace page cache will use all available memory, similar to the OS page cache. In tools like `top` and `ps`, the clickhouse server process will typically show resident set size near 100% of the machine's RAM - this is normal, and most of this memory is actually reclaimable by the OS on memory pressure (`MADV_FREE`). This behavior can be disabled with server setting `page_cache_use_madv_free = 0`, making the userspace page cache just use a fixed amount of memory `page_cache_size` with no special interaction with the OS. On Mac OS, `page_cache_use_madv_free` is always disabled as it doesn't have lazy `MADV_FREE`.
-
-Unfortunately, `page_cache_use_madv_free` makes it difficult to tell if the server is close to running out of memory, since the RSS metric becomes useless. Async metric `UnreclaimableRSS` shows the amount of physical memory used by the server, excluding the memory reclaimable by the OS: `select value from system.asynchronous_metrics where metric = 'UnreclaimableRSS'`. Use it for monitoring instead of RSS. This metric is only available if `page_cache_use_madv_free` is enabled.
-
 ## Storing Data on Web Server {#storing-data-on-webserver}
 
 There is a tool `clickhouse-static-files-uploader`, which prepares a data directory for a given table (`SELECT data_paths FROM system.tables WHERE name = 'table_name'`). For each table you need, you get a directory of files. These files can be uploaded to, for example, a web server with static files. After this preparation, you can load this table into any ClickHouse server via `DiskWeb`.

docs/zh/getting-started/example-datasets/opensky.mdx

@@ -1,4 +1,4 @@
----
+--- 
 slug: /zh/getting-started/example-datasets/opensky
 sidebar_label: 空中交通数据
 description: 该数据集中的数据是从完整的 OpenSky 数据集中衍生而来的，对其中的数据进行了必要的清理，用以展示在 COVID-19 期间空中交通的发展。
@@ -53,12 +53,12 @@ CREATE TABLE opensky
 ls -1 flightlist_*.csv.gz | xargs -P100 -I{} bash -c 'gzip -c -d "{}" | clickhouse-client --date_time_input_format best_effort --query "INSERT INTO opensky FORMAT CSVWithNames"'
 ```
 
-- 这里我们将文件列表（`ls -1 flightlist_*.csv.gz`）传递给`xargs`以进行并行处理。 `xargs -P100` 指定最多使用 100 个并行工作程序，但由于我们只有 30 个文件，工作程序的数量将只有 30 个。
-- 对于每个文件，`xargs` 将通过 `bash -c` 为每个文件运行一个脚本文件。该脚本通过使用 `{}` 表示文件名占位符，然后 `xargs` 由命令进行填充（使用 `-I{}`）。
-- 该脚本会将文件 (`gzip -c -d "{}"`) 解压缩到标准输出（`-c` 参数），并将输出重定向到 `clickhouse-client`。
-- 我们还要求使用扩展解析器解析 [DateTime](/docs/zh/sql-reference/data-types/datetime.md) 字段 ([--date_time_input_format best_effort](/docs/zh/operations/settings/settings.md#settings-date_time_input_format)) 以识别具有时区偏移的 ISO-8601 格式。
+- 这里我们将文件列表（`ls -1 flightlist_*.csv.gz`）传递给`xargs`以进行并行处理。 `xargs -P100` 指定最多使用 100 个并行工作程序，但由于我们只有 30 个文件，工作程序的数量将只有 30 个。 
+- 对于每个文件，`xargs` 将通过 `bash -c` 为每个文件运行一个脚本文件。该脚本通过使用 `{}` 表示文件名占位符，然后 `xargs` 由命令进行填充（使用 `-I{}`）。 
+- 该脚本会将文件 (`gzip -c -d "{}"`) 解压缩到标准输出（`-c` 参数），并将输出重定向到 `clickhouse-client`。 
+- 我们还要求使用扩展解析器解析 [DateTime](../../sql-reference/data-types/datetime.md) 字段 ([--date_time_input_format best_effort](../../operations/settings/ settings.md#settings-date_time_input_format)) 以识别具有时区偏移的 ISO-8601 格式。
 
-最后，`clickhouse-client` 会以 [CSVWithNames](/docs/zh/interfaces/formats.md#csvwithnames) 格式读取输入数据然后执行插入。
+最后，`clickhouse-client` 会以 [CSVWithNames](../../interfaces/formats.md#csvwithnames) 格式读取输入数据然后执行插入。 
 
 并行导入需要 24 秒。
 

programs/server/Server.cpp

@@ -1228,13 +1228,6 @@ try
     }
     global_context->setMarkCache(mark_cache_policy, mark_cache_size, mark_cache_size_ratio);
 
-    size_t page_cache_size = server_settings.page_cache_size;
-    if (page_cache_size != 0)
-        global_context->setPageCache(
-            server_settings.page_cache_chunk_size, server_settings.page_cache_mmap_size,
-            page_cache_size, server_settings.page_cache_use_madv_free,
-            server_settings.page_cache_use_transparent_huge_pages);
-
     String index_uncompressed_cache_policy = server_settings.index_uncompressed_cache_policy;
     size_t index_uncompressed_cache_size = server_settings.index_uncompressed_cache_size;
     double index_uncompressed_cache_size_ratio = server_settings.index_uncompressed_cache_size_ratio;

src/Access/Common/AccessType.h

@@ -162,7 +162,6 @@ enum class AccessType
     M(SYSTEM_DROP_COMPILED_EXPRESSION_CACHE, "SYSTEM DROP COMPILED EXPRESSION, DROP COMPILED EXPRESSION CACHE, DROP COMPILED EXPRESSIONS", GLOBAL, SYSTEM_DROP_CACHE) \
     M(SYSTEM_DROP_FILESYSTEM_CACHE, "SYSTEM DROP FILESYSTEM CACHE, DROP FILESYSTEM CACHE", GLOBAL, SYSTEM_DROP_CACHE) \
     M(SYSTEM_SYNC_FILESYSTEM_CACHE, "SYSTEM REPAIR FILESYSTEM CACHE, REPAIR FILESYSTEM CACHE, SYNC FILESYSTEM CACHE", GLOBAL, SYSTEM) \
-    M(SYSTEM_DROP_PAGE_CACHE, "SYSTEM DROP PAGE CACHE, DROP PAGE CACHE", GLOBAL, SYSTEM_DROP_CACHE) \
     M(SYSTEM_DROP_SCHEMA_CACHE, "SYSTEM DROP SCHEMA CACHE, DROP SCHEMA CACHE", GLOBAL, SYSTEM_DROP_CACHE) \
     M(SYSTEM_DROP_FORMAT_SCHEMA_CACHE, "SYSTEM DROP FORMAT SCHEMA CACHE, DROP FORMAT SCHEMA CACHE", GLOBAL, SYSTEM_DROP_CACHE) \
     M(SYSTEM_DROP_S3_CLIENT_CACHE, "SYSTEM DROP S3 CLIENT, DROP S3 CLIENT CACHE", GLOBAL, SYSTEM_DROP_CACHE) \