Set up multiple hot standby servers in multiple data centers
Our core operating system has set up multiple hot standby servers in multiple data centers to provide the reliability and stability in system configurations. This failover mechanism will ensure nothing will affect our operating system even when one server or one data canter cannot provide normal services for any reason.
A Real-time monitoring system and automatic fault recovery mechanism
To ensure system stability, the real-time monitoring system will monitor the whole system and will have alert. Also, this monitoring system will monitor all the servers and set up automatic fault recovery for each server.
Two dedicated monitoring system will monitor two different data centers (24/7). The monitoring services include:
WEB server port, to monitor externally the activation status and the response speed
Mail server port, to ensure the normal operation of the email service
Domain Name Service port, to ensure the normal operation of the domain name server
Database server, to monitor service internally and to apply automatic recovery mechanism to the service
WEB server, to monitor internally and to apply automatic recovery mechanism to the service
Server endpoint SOCKET connection
A real-time intrusion detection and alarm system
Our operating system provides three distinct levels of security:
Data center level - defend against DDoS attacks, defend against ARP when it attacks firewall
Internet level - real-time monitoring of the firewall
Server level - All the servers have installed the real-time intrusion detection and alarm system software. When the server detects abnormalities such as being attacked, it will first block the attacking source, while sending real-time text messages to notify the relevant system maintenance personnel.
System backup and disaster recovery
Dinpay backup system includes system regular backups as well as backups and recovery of thereal-time operational databases.
System regular backups are mainly responsible for the site data backup, in the way of using multi-server hot standby in different data centers.
Database real-time backups use database cluster. Every data center uses database cluster to process requests of accessing database. When the number of requests increase, it only need to increase the number of node in a database cluster. When one data node is not accessible, all the requests to access database will go to the remote database server cluster.
Every week, each data center will execute the full system backup based on other storage media (large capacity removable memory).When the central database is not accessible due to the disaster, the storage media will deliver the bulk data to the new designated data center and set up a system-wide disaster recovery.