Applications for which distributed SQL is particularly suitable
Posted: Mon Jan 27, 2025 4:38 am
Distributed SQL is recommended for all application scenarios with the following characteristics:
The size of the database reaches the maximum capacity of the RDBMS or the hardware used. A database cluster is helpful here by evenly distributing the entire database across different nodes.
There are a large number of concurrent client connections. Distributed SQL can handle significantly more concurrently open connections by distributing the connections across multiple nodes.
High availability with high fault tolerance is required. This can only be achieved with distributed SQL, since the systems have no primary nodes and each node handles albania telegram screening all the data traffic. This avoids the negative consequences of connection failures to the clients.
The application is very read-intensive - the even distribution of read operations across the cluster ensures that there is no drop in performance.
The application is very write intensive – again, distributed SQL is superior because the writes are evenly distributed across the slices throughout the cluster.
3. Properly configure a cluster for distributed SQL
MariaDB Xpand is the database system for distributed SQL that, thanks to its compatibility with the MariaDB Server, allows a company to easily upgrade from a conventional client-server system to a database cluster. Two application packages are required for deployment: Xpand and MaxScale in the latest releases. Xpand is the engine for distributed SQL and MaxScale is the database proxy - the backend for applications. Installation of these two packages on each cluster server is required.
At this point, an important note on the hardware requirements: A MaxScale node requires an eight-core processor and 16 GB of memory. The individual Xpand nodes require 8 to 32 cores and 64 GB of memory each. Each should also offer at least 20 GiB of SSD storage. The possible number of nodes and processor cores on production systems is specified in the license agreement and tested using the license key.
The size of the database reaches the maximum capacity of the RDBMS or the hardware used. A database cluster is helpful here by evenly distributing the entire database across different nodes.
There are a large number of concurrent client connections. Distributed SQL can handle significantly more concurrently open connections by distributing the connections across multiple nodes.
High availability with high fault tolerance is required. This can only be achieved with distributed SQL, since the systems have no primary nodes and each node handles albania telegram screening all the data traffic. This avoids the negative consequences of connection failures to the clients.
The application is very read-intensive - the even distribution of read operations across the cluster ensures that there is no drop in performance.
The application is very write intensive – again, distributed SQL is superior because the writes are evenly distributed across the slices throughout the cluster.
3. Properly configure a cluster for distributed SQL
MariaDB Xpand is the database system for distributed SQL that, thanks to its compatibility with the MariaDB Server, allows a company to easily upgrade from a conventional client-server system to a database cluster. Two application packages are required for deployment: Xpand and MaxScale in the latest releases. Xpand is the engine for distributed SQL and MaxScale is the database proxy - the backend for applications. Installation of these two packages on each cluster server is required.
At this point, an important note on the hardware requirements: A MaxScale node requires an eight-core processor and 16 GB of memory. The individual Xpand nodes require 8 to 32 cores and 64 GB of memory each. Each should also offer at least 20 GiB of SSD storage. The possible number of nodes and processor cores on production systems is specified in the license agreement and tested using the license key.