General Specifications
- [Rule] Database valuable features
Table 1 Recommended database valuable features Category
Feature List
Description
Table type
DISTRIBUTION table
Data sharding.
PARTITION table
Data partitioning.
Storage engine
Row store
Stores tables by row, which is recommended in the point query and a scenario with many add, delete, and modify operations.
Transaction
Transaction block
Explicitly starts a transaction.
Single-statement transaction
Does not explicitly start a transaction. A single statement is a transaction.
Distributed transaction (weak consistency)
GTM-free mode. Strong consistency can be ensured in the sharding scenario.
Scale-out
Online scale-out
Ensures smooth transition of user services during node scale-out and data redistribution.
Deployment
One primary and multiple standby DNs
Supports DN sharding. One primary DN and multiple standby DNs are used to ensure data backup and disaster recovery.
Security
Transparent data encryption
Supports database-level storage encryption. Upper-layer services are unaware of the encryption.
Data type
Integer
TINYINT, SMALLINT, INTEGER, BIGINT
Arbitrary-precision
NUMERIC/DECIMAL
Floating-point
REAL/FLOAT4,DOUBLE PRECISION/FLOAT8,FLOAT
Boolean
Boolean
Fixed-length character
CHAR(n)
Variable-length character
VARCHAR(n), NVARCHAR2(n), and TEXT
Time
DATE, TIME, TIMETZ, TIMESTAMP, TIMESTAMPTZ, SMALLDATETIME, INTERVAL, and REALTIME
Binary
BYTEA (variable-length binary)
Bit string
BIT(n) and VARBIT(n)
Function
String function
String data type processing function.
Binary string function
Binary string type processing function.
Numeric operation function
Numeric type processing function.
Time and date processing function
Time and date type processing function.
Index
Primary key or unique index
Single-column or multi-column primary key or unique index.
B-tree index
Index type.
- [Rule] Recommended best practices for database usage
Table 2 Recommended best practices for database usage No.
Item
Recommended Value
1
Optimal number of shards in a cluster (number of primary DNs)
< 256
2
Optimal number of long connections in a cluster
For details, see the default configuration of GUC parameter max_connections in the corresponding hardware specifications.
3
Maximum data volume on a single physical node
16 TB (Determine the capacity based on the backup and restoration specifications.)
4
Number of active databases in a cluster
1
5
Total number of cluster tables and indexes
10000 (It is recommended that the number of tables in a single schema be less than or equal to 200.)
NOTE:The value is only a recommended value. You need to plan the value based on the disk capacity and customer services. Theoretically, the maximum value is 232.
6
Optimal number of columns in a single table
< 50
7
Optimal number of indexes in a table
< 5
8
Optimal number of composite indexes in a table
< 3
9
Optimal number of columns in a single composite index
< 5
10
Optimal width of a single row
< 2000
11
Optimal capacity of a single column
< 10 MB
12
Optimal SQL statement length
< 5000
13
Percentage of available disk space
50%
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.See the reply and handling status in My Cloud VOC.
For any further questions, feel free to contact us through the chatbot.
Chatbot