Provisioning the hyper-converged infrastructure bare metal systems from the top of the rack switch
Methods and devices for provisioning a hyper-converged infrastructure of bare metal systems are disclosed herein. Two fabric elements are configured in a master-slave arrangement to ensure high availability. ONIE capable fabric elements may be pre-installed with an operating system as firmware to run open network operating systems, such as Linux. The Linux operating system includes a KVM hypervisor to run virtual machines. An operating system of the virtual machines can access an external network by creating a bridge between switch management ports and a virtual network interface. New node elements may be added by connecting the network ports of the new node element to the fabric elements and booting the new node element in a network/PXE boot mode. The new node element obtains an IP address from a DHCP server and boots an image downloaded from a PXE server.
1. A method for provisioning a node of a hyper-converged infrastructure, the method comprising:
booting the node without an operating system, wherein:
the node comprises an Intelligent Platform Management Interface (IPMI) port and a switch management port;
the IPMI port and the switch management port are coupled to a network switch and a fabric element in the hyper-converged infrastructure; and
the fabric element provides data connectivity between the node and at least one other node element through a direct connection to a high speed data port of the node;
allocating an internet protocol address to the node using a dynamic host configuration protocol (DHCP) server executed by a virtual machine, wherein:
the virtual machine is hosted by the fabric element;
the virtual machine executes a preboot execution environment (PXE); and
the PXE stores pre-built operating system images;
downloading an operating system image from the PXE of the virtual machine to the node;
installing an operating system to a local disk of the node using the operating system image; and
booting the node using the operating system.
2. The method of claim 1 , wherein the fabric element is pre-installed with a second operating system as firmware configured to execute an open network operating system.
3. The method of claim 2 , wherein:
the firmware is configured to execute a Linux operating system; and
the fabric element executes a kernel-based virtual machine hypervisor using the virtual machine.
4. The method of claim 2 , wherein:
the fabric element is configured as a master fabric element;
the hyper-converged infrastructure further comprises a second fabric element configured as a slave fabric element; and
the fabric element and the second fabric element interoperate to improve availability of the hyper-converged infrastructure.
5. The method of claim 1 , wherein the switch management port is a 1G port.
6. The method of claim 1 , further comprising configuring the DHCP server with a range of IP addresses for allocation.
7. The method of claim 1 , wherein the fabric element is a top of the rack switch that controls a board management controller of the node using the IPMI port.
8. The method of claim 1 , further comprising accessing an external network by creating a bridge between the switch management port and a virtual network interface of the virtual machine.
9. A system comprising:
a stateless node of a hyper-converged infrastructure;
means for booting the stateless node without an operating system, wherein:
the stateless node comprises a main memory and an Intelligent Platform Management Interface (IPMI) port coupled to a network switch and a fabric element in the hyper-converged infrastructure; and
the fabric element is configured to provide connectivity between the node and at least one other node element through a direct connection to a high speed data port of the node; and
means for allocating an internet protocol address to the stateless node using a dynamic host configuration protocol (DHCP) server executed by a virtual machine, wherein:
the virtual machine is hosted by the fabric element;
the virtual machine is configured to execute a preboot execution environment (PXE); and
the PXE stores pre-built operating system images;
means for obtaining bootstrap software at the stateless node from the fabric element; and
means for executing the bootstrap software in the main memory to boot the stateless node.
10. The system of claim 9 , wherein:
the fabric element is configured as a master fabric element;
the hyper-converged infrastructure further comprises a second fabric element configured as a slave fabric element; and
the fabric element and the second fabric element are configured to interoperate to improve availability of the hyper-converged infrastructure.
11. The system of claim 9 , wherein:
the stateless node further comprises a switch management port coupled to the network switch; and
the switch management port is a 1G port.
12. The system of claim 9 , further comprising means for configuring the DHCP server with a range of IP addresses for allocation.
13. The system of claim 9 , wherein the fabric element is a top of the rack switch that is configured to control a board management controller of the stateless node using the IPMI port.
14. The system of claim 9 , further comprising means for accessing an external network by creating a bridge between a switch management port of the stateless node and a virtual network interface of the virtual machine.
15. A device for provisioning a hyper-converged infrastructure, the device comprising:
a fabric element configured to run a virtual machine and coupled to a network switch, wherein:
the fabric element is configured to provide connectivity among a plurality of nodes through a direct connection to high speed data ports of the plurality of nodes;
the virtual machine is configured to execute a preboot execution environment (PXE); and
the PXE is configured to store pre-built operating system images; and
a node, from among the plurality of nodes, coupled to the fabric element, wherein the node comprises an Intelligent Platform Management Interface (IPMI) port and a switch management port, the IPMI port and the switch management port are configured to be coupled to the network switch, and the node is configured to:
download an operating system image from the PXE of the virtual machine;
install an operating system to a local disk of the node using the operating system image; and
boot using the operating system.
16. The device of claim 15 , wherein the fabric element and the node comprise high speed data ports for coupling the fabric element to the node.
17. The device of claim 15 , wherein:
the virtual machine is configured to execute a dynamic host configuration protocol (DHCP) server; and
the node is further configured to obtain an internet protocol address from the DHCP server.
18. The device of claim 15 , wherein the fabric element is configured as a master fabric element, and further comprising a second fabric element configured as a slave fabric element and configured to be coupled to the network switch and the node, wherein the fabric element and the second fabric element are configured to interoperate to improve availability of the hyper-converged infrastructure.
19. The device of claim 15 , wherein the fabric element is pre-installed with a second operating system as firmware configured to execute an open network operating system.
20. The device of claim 19 , wherein:
the firmware is configured to execute a Linux operating system; and
the fabric element is configured to execute a kernel-based virtual machine hypervisor using the virtual machine.