Method and apparatus for providing a remotely managed expandable computer system

• US 20110153798A1
• Filed: 12/22/2009
• Published: 06/23/2011
• Est. Priority Date: 12/22/2009
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Dormitzer, Paul, Carrigan, Douglas, Gorius, Aaron, Wilde, Myles, Fallon, Michael, Wheeler, William, Groenendaal, Johan van de, Yang, Chengda, Adiletta, Matthew
• Granted Patent: US 8,667,110 B2
• Time in Patent Office: Days
• US Class Current: 709/223
• CPC Class Codes: H04L 63/0428 wherein the data content is protected, e.g. by encrypting or encapsulating the payload
• CPC Class Codes: H04L 63/08 for authentication of entities cryptographic mechanisms or cryptographic arrangements for entity authentication H04L9/32
• CPC Class Codes: H04L 63/20 for managing network security; network security policies in general filtering policies H04L63/0227

Abstract

A remotely managed expandable multi-user computer system is provided. The computer system includes a removable compute cartridge and a removable storage cartridge allowing easy upgrade of the computer system. The computer system self-configures when powered on to allow remote management by another system over a network. Automatic failover support is provided without the need for a separate compute element. A self-configurable removable secure-to-wireless converter to couple to a client computer system allows secure communications between the multi-user system and the client computer system via a wireless network.

Claims

1. An apparatus comprising:

a backplane including a plurality of slots;

a replaceable stateless storage cartridge; and

at least two replaceable stateless compute cartridges, each compute cartridge comprising;

an interface to a communications network, the compute cartridge remotely managed via the communications network; and

memory to store a distributed manager, the distributed manager to dynamically maximize compute resource utilization over the plurality of compute cartridges.

2. The apparatus of claim 1, wherein each slot in the backplane is adapted to receive either the storage cartridge or the compute cartridge.

3. The apparatus of claim 1, wherein appliance services stored in a remote system are downloaded by the remote system to be stored in the memory for immediate use by the compute cartridge.

4. The apparatus of claim 1, wherein a virtual resource manager for a first compute cartridge is selected to be a master virtual resource manager and the other virtual resource manager for the other compute cartridge is selected to be a slave virtual resource manager, the master virtual resource manager to be responsible for global management of virtual machines.

5. The apparatus of claim 4, further comprising:

a dedicated management network, the dedicated management network used by the plurality of virtual resource managers to communicate securely between compute cartridges.

6. The apparatus of claim 4, wherein each slave virtual resource manager including a registration number, the registration number representing an order in which the slave virtual resource manager has registered with the master resource manager, each registration number associated with a different failover timeout period.

7. The apparatus of claim 6, wherein upon detecting a failure to receive a beacon from a master virtual resource manager after a timeout interval, the slave virtual resource manager having the lowest registration number to transparently transition to master virtual resource manager.

8. The apparatus of claim 4, wherein a resiliency software layer within the distributed manager automatically ensures that a manager virtual resource manager is operating.

9. The apparatus of claim 1, wherein the memory to store a distributed virtual manager configuration data structure that associates data on a storage device in a storage cartridge with a configuration.

10. A method comprising:

dynamically maximizing compute resource utilization over at least two replaceable stateless compute cartridges installed in a respective slot of a backplane of a computer system, each compute cartridge including a memory to store a distributed manager; and

remotely managing the computer system, by a remote system via a communications network through an interface to the communications network in the compute cartridge.

11. The method of claim 10, wherein each slot in the backplane is adapted to receive either the storage cartridge or the compute cartridge.

12. The method of claim 10, further comprising:

receiving appliance services stored in a remote system, the appliance services downloaded by the remote system to be stored in memory in the compute cartridge for immediate use by the compute cartridge.

13. The method of claim 12, further comprising:

selecting a virtual resource manager for a first compute cartridge as a master virtual resource manager; and

selecting the other virtual resource manager for the other compute cartridge as a slave virtual resource manager, the master virtual resource manager responsible for global management of virtual machines.

14. The method of claim 13, further comprising:

providing a dedicated management network between compute cartridges to allow the plurality of virtual resource managers to communicate securely.

15. The method of claim 13, wherein each slave virtual resource manager includes a registration number, the registration number representing an order in which the slave virtual resource manager has registered with the master resource manager, each registration number associated with a different failover timeout period.

16. The method of claim 15, wherein upon detecting a failure to receive a beacon from a master virtual resource manager after a timeout interval, the slave virtual resource manager having the lowest registration number to transparently transition to master virtual resource manager.

17. The method of claim 13, wherein a resiliency software layer within the distributed manager automatically ensures that a manager virtual resource manager is operating.

18. The method of claim 10, wherein the memory to store a distributed virtual manager configuration data structure that associates data on a storage device in a storage cartridge with a configuration.

19. A system comprising:

a multi-user system, the multi-user system comprising a

a replaceable stateless storage cartridge including a storage device, at least two replaceable stateless compute cartridges, and a wireless communications port;

a client system, the client system comprising a wired network port; and

a converter removeably coupled to the wired network port in the client system, the converter comprising;

a converter wireless interface to communicate via a wireless network using a wireless communications protocol;

a converter wired interface to communicate via a wired network using a wired communications protocol;

a non-volatile memory to store a unique identifier, the unique identifier stored in the non-volatile memory by a manufacturer and not configurable by a user, the unique identifier used to provide secure communication over the wireless network; and

a packet converter to perform packet conversion between the wireless communications protocol and the wired communications protocol, the converter wireless interface to automatically provide secure communication using the stored unique identifier to encrypt and decrypt all packets transferred over the wireless network between the client system and the multi-user system.

20. The system of claim 19, wherein each slot in the backplane is adapted to receive either the storage cartridge or the compute cartridge

21. The system of claim 19, wherein the packet converter adds authentication information to packets prior to sending packets over the wireless network.

22. The system of claim 21, wherein the packet converter verifies authenticity of packets received over the wireless network prior to sending the packets over the wired network.