Table Configuration

The table configurations are listed below:

Configurations	Description	Options	Updated
number_of_shards	Number of shards	Greater than 1, defaults to 0	No
number_of_replicas	Number of replicas	[1, 5], defaults to 1	Yes
create_date	Table creation date		No
data_center	Cross-datacenter related configurations. Represents primary data center	Defaults to “DEFAULT”	Yes
disaster_preparedness	Cross-datacenter related configurations. Represents if high availability across data centers is enabled	True/False, defaults to false	Yes
scatter_replica	Cross-datacenter related configurations. Generally, at least 1 replica will be allocated to each data center. When set to true, replicas will be allocated to all data centers as much as possible. If there are 5 replicas and 3 data centers, 1 center will have 1 replica, and the others will have 2 for each of them. When set to false, replicas will be allocated mainly to primary data center. Thus if there are 5 replicas and 3 data centers, primary data center will have 3 replicas, and other two centers will have one for each	True/False, defaults to false	Yes
dc_affinity	Cross-datacenter related configurations. When set to True, Hippo will try to migrate the leadership of replicas into primary data center	True/False, defaults to false	Yes
tag	Tag		Yes
embedding.segment.max_size_mb	Expected size of each index segment	Defaults to 512 MB	Yes
embedding.segment.seal_proportion	When the size of Flat index in the memory exceeds a certain percentage of the value set in “embedding.segment.max_size_mb”, Hippo will generate vector index automatically for incremental data. This parameter is used for configuring this percentage	Defaults to 0.2	Yes
embedding.segment.max_deletion_proportion	When the volume of data deleted in index segments exceeds a certain percentage, Hippo will automatically compact vector index segments to release available resources. This parameter is used for configuring this percentage	Defaults to 0.1	Yes

Table 12 Table Configurations

Check Table Configuration

Users can use the following commands to check the table configuration.

curl -u shiva:shiva -XGET 'localhost:8902/hippo/v1/_settings?database_name={database_name}&pretty'
curl -u shiva:shiva -XGET 'localhost:8902/hippo/v1/{table}/_settings?database_name={database_name}&pretty'
curl -u shiva:shiva -XGET 'localhost:8902/hippo/v1/{table}/_setting?database_name={database_name}&pretty'

Result:

{
  "default#book" : {
    "number_of_shards" : 1,
    "creation_date" : "2023-06-15 11:42:25",
    "number_of_replicas" : 1,
    "data_center" : "DEFAULT",
    "dc_affinity" : false,
    "disaster_preparedness" : false,
    "scatter_replica" : false
  }
}

Parameter description:

Parameters	Description	Required
table	Table to be matched, supports using asterisk (*) wildcard character and specifying multiple pattens, which are separated by comma (,)	No, defaults to *, gets all table information of current database
database_name	The database where the table is located	No, defaults to "default" database

Table 13 Check Table Configuration

Update Table Configuration

Users can use the following commands to update table configuration.

curl -u shiva:shiva -XPUT 'localhost:8902/hippo/v1/{table}/_settings?database_name={database_name}&pretty' -H 'Content-Type: application/json' -d'{
  "number_of_replicas" : 1,
  "data_center" : "beijing",
  "disaster_preparedness" : true,
  "scatter_replica" : true,
  "dc_affinity" : true,
  "tag" : "newTag",
  "tag.clear" : true,
  "embedding.segment.max_size_mb" : 512,
  "embedding.segment.seal_proportion" : 0.2,
  "embedding.segment.max_deletion_proportion" : 0.1
}';

Result:

{
  "acknowledged" : true
}

Parameter description:

Parameters	Description	Required
table	Table to be matched, supports using asterisk (*) wildcard character and specifying multiple pattens, which are separated by comma (,)	Yes
database_name	The database where the table is located	No, defaults to "default" database

Table 14 Update Table Configuration