The new architecture, known as Global Scale, is claimed to allow for greater scalability, better cost-efficiency, and global distribution.
"Nextcloud Global Scale works by effectively removing the need for shared components in the existing architecture like the load balancers, hosting centre uplink, database, storage, and cache," the company said in a statement.
"It uses multiple independent application servers, called nodes, each running on standard, inexpensive commodity hardware. Storage, database, and cache are running local on the application servers and no longer have to be kept in sync."
"By automatically moving user data between nodes to achieve an optimal balance between legal requirements, costs and Quality of Service, GS removes traditional, costly scalability limitations like data centre uplink, load balancers and database, storage and cache."
Nextcloud Global Scale is claimed to eliminate the need for a central database, storage and cache solution, and enable individual nodes to take advantage of commodity hardware to save costs and increase flexibility. It has three main components:
The nodes are accessed by users through the Global Site Selector which redirect the user to the node their data currently resides on;
Sharing is mediated through the Lookup Server (LS) which is queried by internal or external share requests and resolves the current location of the shared data. The LS also is responsible for tracking storage and quota settings as well as Quality of Service metrics like speed and reliability class and the physical location of users.
The final component of GS is the Balancer which is responsible for monitoring the load of individual nodes and initiating the migration of users between nodes based on QoS settings and changes in physical location of users.