System and methods for load placement in data centers
A system for operating information handling systems forming a network including a plurality of switches is provided. The system includes an open flow controller coupled to each of the plurality of switches; a plurality of links, each link configured to transmit data packets between two switches from the plurality of switches; wherein: the open flow controller is configured to determine a traffic flow across each of the plurality of links; and each one of the plurality of switches is configured to re-route a data packet when the traffic flow in a link associated to the switch exceeds a threshold. A computer program product including a non-transitory computer readable medium having computer readable and executable code for instructing a processor in a management unit for a plurality of information handling systems as above is also provided. A network managing device coupled to a service provider having resources is also provided.
1. A network managing device comprising:
one or more network ports coupling the network managing device to a plurality of switches;
a memory storing commands and data; and
a processing circuit coupled to the memory and the one or more network ports, the processing circuit being configured to:
discover a topology of a network comprising the plurality of switches;
determine a path for network traffic between a source and a destination based on the topology;
determine a fail re-route (FRR) path set for a first switch from the plurality of switches along the path, the FRR path set including a point of failure specific FRR path for each potential point of failure along the path between the first switch and the destination, each of the FRR paths in the FRR path set including an instruction to the first switch to perform a push, swap, or pop label operation; and
provide the FRR path set to the first switch for use by the first switch to reroute the network traffic when a failure along the path is detected by the first switch and without the first switch involving the network managing device when the failure is detected.
2. The network managing device of claim 1 , wherein the failure is a link failure.
3. The network managing device of claim 1 , wherein the failure is a switch failure.
4. The network managing device of claim 1 , wherein the failure is a combination of a switch failure and a link failure.
5. The network managing device of claim 1 , wherein the network managing device is an open flow controller.
6. The network managing device of claim 1 , wherein the processing circuit is further configured to receive notification of the failure from the first switch.
7. The network managing device of claim 1 , wherein the processing circuit is further configured to:
determine traffic rates on a first link along the path; and
in response to determining that the traffic rates on the first link exceed a threshold, send one or more commands to one or more of the plurality of switches to direct network traffic away from the first link.
8. The network managing device of claim 7 , wherein the one or more commands comprise a command to one of the plurality of switches to perform a push label operation or a swap label operation.
9. The network managing device of claim 1 , wherein:
the path is a label switch path; and
the processing circuit is further configured to:
send a first command to a forwarding switch to perform a push label operation;
send a second command to a receiving switch to perform a pop label operation; and
send one or more third commands to one or more second switches to perform a swap label operation.
10. The network managing device of claim 1 , wherein the processing circuit is further configured to determine the path and the FRR path set in response to detecting a start of a flow.
11. A switch comprising:
one or more network ports coupling the switch to a network managing device and a plurality of other switches;
a memory storing commands and data; and
a processing circuit coupled to the memory and the one or more network ports, the processing circuit being configured to:
receive, from the network managing device, instructions for forwarding network traffic along a path between a source and a destination;
receive, from the network managing device, a fail re-route (FRR) path set, the FRR path set including a point of failure specific FRR path for each potential point of failure along the path between the switch and the destination, each failure specific FRR path including an instruction to the switch to perform a push, swap, or pop label operation;
exchange hand-shake packets with one or more of the other switches along the path between the switch and the destination;
detect a failure along the path between the switch and the destination; and
reroute the network traffic to the destination along a FRR path from the FRR path set that corresponds to the detected failure without involving the network managing device.
12. The switch of claim 11 , wherein the failure is a link failure, a switch failure, or a combination of a switch failure and a link failure.
13. The switch of claim 11 , wherein the network managing device is an open flow controller.
14. The switch of claim 11 , wherein the processing circuit is further configured to report the failure to the network managing device.
15. The switch of claim 11 , wherein the instructions for forwarding network traffic along the path comprise an instruction to perform a push label operation, a swap label operation, or a pop label operation.
16. The switch of claim 11 , wherein the hand-shake packets are bidirectional forwarding detection (BFD) packets.
17. An information handling system comprising:
a network managing device comprising one or more first network ports coupling the network managing device to a plurality of first switches, a first memory storing commands and data, and a first processing circuit coupled to the first memory and the one or more first network ports; and
a second switch comprising one or more second network ports coupling the second switch to the network managing device and the plurality of first switches, a second memory storing commands and data, and a second processing circuit coupled to the second memory and the one or more second network ports;
wherein the network managing device is configured to:
discover a topology of a network comprising the plurality of first switches and the second switch;
determine a path for network traffic between a source and a destination based on the topology, the second switch being located on the path;
determine, in response to detecting a start of a flow, a fail re-route (FRR) path set for the second switch, the FRR path set including a point of failure specific FRR path for each potential point of failure along the path between the second switch and the destination; and
provide the FRR path set to the second switch;
wherein the second switch is configured to:
exchange hand-shake packets with one or more of the first switches along the path between the second switch and the destination;
detect a failure along the path between the second switch and the destination; and
reroute the network traffic to the destination along a FRR path from the FRR path set that corresponds to the detected failure without involving the network managing device.
18. The information handling system of claim 17 , wherein the failure is a link failure, a switch failure, or a combination of a switch failure and a link failure.
19. The information handling system of claim 17 , wherein the network managing device is an open flow controller.
20. The information handling system of claim 17 , wherein the second switch is further configured to report the failure to the network managing device.