This commit is contained in:
sakulali 2024-08-24 00:48:40 +08:00
parent d6731aba8b
commit 67d80bafd2
2 changed files with 2 additions and 2 deletions

View File

@ -252,7 +252,7 @@ This format is also available under the names `TSVRawWithNames`, `RawWithNames`.
Differs from `TabSeparatedWithNamesAndTypes` format in that the rows are written without escaping.
When parsing with this format, tabs or linefeeds are not allowed in each field.
This format is also available under the names `TSVRawWithNamesAndNames`, `RawWithNamesAndNames`.
This format is also available under the names `TSVRawWithNamesAndTypes`, `RawWithNamesAndTypes`.
## Template {#format-template}

View File

@ -755,7 +755,7 @@ Used for the same purpose as `max_block_size`, but it sets the recommended block
However, the block size cannot be more than `max_block_size` rows.
By default: 1,000,000. It works when reading from MergeTree engines and reading contents from row input formats.
## preferred_max_column_in_block_size_bytes {#preferred-max-column-in-block-size_bytes}
## preferred_max_column_in_block_size_bytes {#preferred-max-column-in-block-size-bytes}
Limit on max column size in block while reading. Helps to decrease cache misses count. Should be close to L2 cache size.
However, the column size in block cannot be more than `max_block_size` rows.