Multi-state status reporting for high-availability cluster nodes
View Patent ↗One embodiment disclosed relates to a node system of a high-availability cluster. The node system includes at least a first register and an output port. The first register stores multi-state status data of the node, and the output port sends signals representing this multi-state status data. The multi-state status data includes at least one degraded state. The node system may also include a second register and an input port. The input port receives signals representing the multi-state status data of another node. The second stores this multi-state status data from the other node. Another embodiment disclosed relates to a method of status reporting for a node of a cluster. A set of rules is applied to determine current multi-state status of the node. The states of the multi-state status including a good state, a bad state, and at least one degraded state.
1. A method of status reporting for a computer system configured as a node of a cluster of interconnected computer systems that is used as a single computing unit, the method comprising applying a set of rules to determine current multi-state status of the node, wherein states of the multi-state status includes a good state, a bad state, and at least one degraded state,
wherein the good state requires that both an application and an operating system report an up status and that no severe or critical chassis codes are received from the computer system,
wherein the bad state requires that either an application or an operating system report a down status, or a critical chassis code is received from the computer system, and
wherein a degraded state requires that either the computer system loses greater than a predetermined percentage of performance, or a severe chassis code is received from the computer system.
2. The method of claim 1 , further comprising:
writing the multi-state status of the node to a first register.
3. The method of claim 2 , further comprising:
driving the multi-state status from the first register to a next node via a point-to-point communications path.
4. The method of claim 3 , further comprising:
receiving multi-state status from another node; and
writing the multi-state status from the other node to a second register.
5. The method of claim 4 further comprising:
reading the statuses from the first and second registers; and
taking action to maintain high availability of the cluster based on the statuses read.
6. The method of claim 4 , wherein the multi-state status from the other node includes multiple levels of degradation.
7. The method of claim 3 wherein the status writable into the second register includes a no signal state.
8. The method of claim 1 wherein the multi-state status of the node includes multiple levels of degradation.
9. The method of claim 1 wherein the set of rules includes a rule such that receipt of a chassis code of a critical level results in the bad state and another rule such that receipt of a chassis code of a level below critical results in a degraded state.