System and method for failure detection in rings

1. A switch in a ring of switches, the switch comprising:

a control unit;

one or more first ports coupled to the control unit, each of the one or more first ports being configured to directly couple the switch to a first neighboring switch in the ring of switches; and

one or more second ports coupled to the control unit, each of the one or more second ports being configured to directly couple the switch to a second neighboring switch in the ring of switches;

wherein the control unit is configured to:

transmit a first request message using one of the one or more first ports, the first request message including a first bitmap that includes a first plurality of bits representing each switch in the ring of switches, each of the first plurality of bits being initialized to an inactive state;

transmit a second request message using one of the one or more second ports, the second request message including a second bitmap that includes a second plurality of bits representing each switch in the ring of switches, each of the second plurality of bits being initialized to the inactive state;

receive a first response message on one of the one or more first ports, the first response message including the first bitmap, the first bitmap having been altered by switching one or more of the first plurality of bits to an active state to indicate each of one or more first switches in the ring of switches that received the first request message;

receive a second response message on one of the one or more second ports, the second response message including the second bitmap, the second bitmap having been altered by switching one or more of the second plurality of bits to the active state to indicate each of one or more second switches in the ring of switches that received the second request message;

detect a failure in the ring of switches in response to receiving the first response message and the second response message;

extract the first bitmap from the first response message;

determine a first reachability of each switch in the ring of switches via the one or more first ports by examining which of the first plurality of bits in the first bitmap are in the active state;

extract the second bitmap from the second response message;

determine a second reachability of each switch in the ring of switches via the one or more second ports by examining which of the second plurality of bits in the second bitmap are in the active state; and

determine whether all of the ring of switches is reachable based on the first reachability and the second reachability.

2. The switch of claim 1 wherein the first request message further includes a count of a number of the one or more first switches.

3. The switch of claim 1 wherein logic one corresponds to the active state.

4. The switch of claim 1 wherein the first request message further includes a route trace entry for each of the one or more first switches.

5. The switch of claim 1 , wherein the control unit is further configured to notify one or more devices selected from a group consisting of a master switch and a network controller of the failure.

6. The switch of claim 1 , wherein the control unit is further configured to transmit one or more error recovery messages to the ring of switches.

7. The switch of claim 1 , wherein the control unit is further configured to:

receive a third request message including a third bitmap on a third port, the third bitmap including a third plurality of bits representing each switch in the ring of switches, wherein a bit among the third plurality of bits that represents the switch is initialized to an inactive state;

alter the third request message to update the third bitmap by setting the bit that represents the switch to the active state;

determine whether the altered third request message can be forwarded around the ring of switches;

when the altered third request message can be forwarded around the ring of switches, forward the altered third request message; and

when the altered third request message cannot be forwarded around the ring of switches:

convert the altered third request message to a third response message; and

transmit the third response message on a fourth port.

8. The switch of claim 7 , wherein the third port and the fourth port are included in the one or more first ports.

9. The switch of claim 7 , wherein the third port and the fourth port are included in the one or more second ports.

10. The switch of claim 1 , wherein the control unit is further configured to:

receive a third response message; and

forward the third response message around the ring of switches.

11. A method of detecting a failure in a ring of switches, the method comprising:

transmitting a first request message using one of one or more first ports of a first switch, each of the one or more first ports being configured to directly couple the first switch to a first neighboring switch in the ring of switches, the first request message including a first bitmap that includes a first plurality of bits representing each switch in the ring of switches, each of the first plurality of bits being initialized to an inactive state;

transmitting a second request message using one of one or more second ports of the first switch, each of the one or more second ports being configured to directly couple the first switch to a second neighboring switch in the ring of switches, the second request message including a second bitmap that includes a second plurality of bits representing each switch in the ring of switches, each of the second plurality of bits being initialized to the inactive state;

receiving a first response message on one of the one or more first ports, the first response message including the first bitmap, the first bitmap having been altered by switching one or more of the first plurality of bits to an active state to indicate each of one or more first switches in the ring of switches that received the first request message;

receiving a second response message on one of the one or more second ports, the second response message including the second bitmap, the second bitmap having been altered by switching one or more of the second plurality of bits to the active state to indicate each of one or more second switches in the ring of switches that received the second request message;

extracting the first bitmap from the first response message;

determining a first reachability of each switch in the ring of switches via the one or more first ports by examining which of the first plurality of bits in the first bitmap are in the active state;

extracting the second bitmap from the second response message;

determining a second reachability of each switch in the ring of switches via the one or more second ports by examining which of the second plurality of bits in the second bitmap are in the active state; and

determine whether all of the ring of switches is reachable based on the first reachability and the second reachability.

12. The method of claim 11 , further comprising notifying one or more devices selected from a group consisting of a master switch and a network controller of the failure.

13. The method of claim 11 , further comprising, transmitting one or more error recovery messages to the ring of switches.

14. An information handling system comprising:

a ring of switches comprising a first switch and a second switch;

the first switch comprising:

a first control unit;

one or more first ports coupled to the first control unit, each of the one or more first ports being configured to directly couple the first switch to a first neighboring switch in the ring of switches; and

one or more second ports coupled to the first control unit, each of the one or more second ports being configured to directly couple the first switch to a second neighboring switch in the ring of switches; the second switch comprising:

a second control unit;

one or more third ports coupled to the second control unit, each of the one or more third ports being configured to couple the second switch to the first switch via one or more switches in the ring of switches including the first neighboring switch; and

one or more fourth ports coupled to the second control unit, each of the one or more fourth ports being configured to couple the second switch to the first switch via one or more switches in the ring of switches including the second neighboring switch;

wherein the first control unit is configured to:

detect a failure in the ring of switches;

transmit a first request message using one of the one or more first ports, the first request message including a first bitmap that includes a first plurality of bits representing each switch in the ring of switches, each of the first plurality of bits being initialized to an inactive state;

transmit a second request message using one of the one or more second ports, the second request message including a second bitmap that includes a second plurality of bits representing each switch in the ring of switches, each of the second plurality of bits being initialized to the inactive state;

receive a first response message on one of the one or more first ports, the first response message including the first bitmap, the first bitmap having been altered by switching one or more of the first plurality of bits to an active state to indicate each of one or more first switches in the ring of switches that received the first request message;

receive a second response message on one of the one or more second ports, the second response message including the second bitmap, the second bitmap having been altered by switching one or more of the second plurality of bits to an active state to indicate each of one or more second switches in the ring of switches that received the second request message;

extract the first bitmap from the first response message;

determine a first reachability of each switch in the ring of switches via the one or more first ports by examining which of the first plurality of bits in the first bitmap are in the active state;

extract the second bitmap from the second response message;

determine a second reachability of each switch in the ring of switches via the one or more second ports by examining which of the second plurality of bits in the second bitmap are in the active state; and

determine whether all of the ring of switches is reachable based on the first reachability and the second reachability;

wherein the second control unit is configured to:

receive a third request message including a third bitmap on one of the one or more fourth ports, the third bitmap including a third plurality of bits representing each switch in the ring of switches, wherein a bit among the third plurality of bits that represents the second switch is initialized to an inactive state;

alter the third request message to update the third bitmap by setting the bit that represents the second switch to the active state;

determine whether the altered third request message can be forwarded on one of the one or more third ports;

when the altered third request message can be forwarded on one of the one or more third ports, forward the altered third request message on one of the one or more third ports;

when the altered third request message cannot be forwarded on one of the one or more third ports:

convert the altered third request message to the first response message; and

transmit the first response message on one of the one or more fourth ports;

receive a third response message on one of the one or more third ports; and

forward the third response message on one of the one or more fourth ports.