Hyper-Converged
Ring Gateway
Servers on networks A and B back each other up. When a primary server on network A fails, business data moves to a standby server on network B, and moves back once the primary network recovers. Plug and play, across domains and networks.
10 + 10
Servers on both networks joined at the same time
10
Servers per network A box
4
Network A boxes per network B box
3
Steps in the switchover
Switchover sequence
Address translation: network B servers can use either network B or network A subnet IPs. Pairings can be added, removed, changed, cleared, and switched on or off.
Specifications
- 10 servers on A and 10 on B joined at the same time
- One A box connects up to 10 A servers
- One B box connects up to 4 A boxes
- Pairings can be added, removed, edited, cleared and switched on or off
Acceptance method (contract criteria)
| Network box | API · Shell · SSH |
|---|---|
| Probe management / box monitoring / configuration | API |
| Interface functions | Configuration interface |
| System image | Delivered as a built system image file |
| Containers | Compatible with common container deployments |
Second-generation gateway
Lower layers rebuilt with layer-3 switching integrated; the switch operating system is compiled from Linux kernel 5.4+ source and supports domestic Chinese CPUs.
- Primary/standby
- Any two networks back each other up, up to 8 A boxes; A and B boxes are themselves redundant
- Security
- MQTT over TLS + built-in CA verification; bidirectional layer-3 hardware encryption of forwarded data; encrypted file system, program bound to chip ID
- Services
- Disaster recovery backup, same-IP masking and takeover, NAT traversal, compatible with public IPs and private clouds
- Management
- API, command line with running-configuration sync, dongle licensing, network diagnostic tools
Cross-network failover projects
Tell us the network topology, number of servers, switchover time and acceptance method.