[doc][fix]Update the description of the doris In_memory attribute to avoid users from interpreting it as an in-memory table (#9892)

Update the description of the doris In_memory attribute to avoid users from interpreting it as an in-memory table
This commit is contained in:
jiafeng.zhang
2022-06-01 08:47:56 +08:00
committed by GitHub
parent 2e0d05e0dc
commit 03a1eec3a5
2 changed files with 13 additions and 9 deletions

View File

@ -271,7 +271,9 @@ distribution_info
* `in_memory`
Use this property to set whether the table is [Memory Table] (DORIS/Operation Manual/Memory Table.md).
Doris has no concept of memory tables.
When this property is set to `true`, Doris will try to cache the data blocks of the table in the PageCache of the storage engine, which has reduced disk IO. But this property does not guarantee that the data block is resident in memory, it is only used as a best-effort identification.
`"in_memory" = "true"`
@ -286,7 +288,7 @@ distribution_info
* Dynamic partition related
The relevant parameters of dynamic partition are as follows:
* `dynamic_partition.enable`: Used to specify whether the dynamic partition function at the table level is enabled. The default is true.
* `dynamic_partition.time_unit:` is used to specify the time unit for dynamically adding partitions, which can be selected as DAY (day), WEEK (week), MONTH (month), HOUR (hour).
* `dynamic_partition.start`: Used to specify how many partitions to delete forward. The value must be less than 0. The default is Integer.MIN_VALUE.
@ -296,7 +298,7 @@ distribution_info
* `dynamic_partition.create_history_partition`: Whether to create a history partition.
* `dynamic_partition.history_partition_num`: Specify the number of historical partitions to be created.
* `dynamic_partition.reserved_history_periods`: Used to specify the range of reserved history periods.
* Data Sort Info
The relevant parameters of data sort info are as follows:
@ -562,6 +564,6 @@ Users can create indexes on multiple columns while building a table. Indexes can
If you add an index in the subsequent use process, if there is data in the table, you need to rewrite all the data, so the creation time of the index depends on the current data volume.
#### Memory table
#### in_memory property
The `"in_memory" = "true"` attribute was specified when the table was created. Doris will try to cache the data blocks of the table in the PageCache of the storage engine, which has reduced disk IO. However, this attribute does not guarantee that the data block is permanently resident in memory, and is only used as a best-effort identification.

View File

@ -270,9 +270,11 @@ distribution_info
`"bloom_filter_columns" = "k1, k2, k3"`
* `in_memory`
* `in_memory`
通过此属性设置该表是否为内存表
Doris 是没有内存表的概念
这个属性设置成 `true`, Doris 会尽量将该表的数据块缓存在存储引擎的 PageCache 中,已减少磁盘IO。但这个属性不会保证数据块常驻在内存中,仅作为一种尽力而为的标识。
`"in_memory" = "true"`
@ -287,7 +289,7 @@ distribution_info
* 动态分区相关
动态分区相关参数如下:
* `dynamic_partition.enable`: 用于指定表级别的动态分区功能是否开启。默认为 true。
* `dynamic_partition.time_unit:` 用于指定动态添加分区的时间单位,可选择为DAY(天),WEEK(周),MONTH(月),HOUR(时)。
* `dynamic_partition.start`: 用于指定向前删除多少个分区。值必须小于0。默认为 Integer.MIN_VALUE。
@ -301,7 +303,7 @@ distribution_info
* 数据排序相关
数据排序相关参数如下:
* `data_sort.sort_type`: 数据排序使用的方法,目前支持两种:lexical/z-order,默认是lexical
* `data_sort.col_num`: 数据排序使用的列数,取最前面几列,不能超过总的key 列数
### Example
@ -564,6 +566,6 @@ Doris 中的表可以分为分区表和无分区的表。这个属性在建表
如果在之后的使用过程中添加索引,如果表中已有数据,则需要重写所有数据,因此索引的创建时间取决于当前数据量。
#### 内存表
#### in_memory 属性
当建表时指定了 `"in_memory" = "true"` 属性。则 Doris 会尽量将该表的数据块缓存在存储引擎的 PageCache 中,已减少磁盘IO。但这个属性不会保证数据块常驻在内存中,仅作为一种尽力而为的标识。