Fallover policy management in high availability systems

A cluster of information handling systems (IHSs), comprising:

• a first IHS including a first processor, a first memory, a first high availability cluster manager (HACM), and first and second applications in the first memory;
• a second IHS including a second processor, a second memory, and a second high availability cluster manager (HACM);
• a network bus coupling the first and second IHSs together;
• and
• a shared storage, that communicates with the first and second IHSs to store IHS state information therefrom, the shared storage communicating with the first and second HACMs;
• the first IHS being configured to dump first IHS state information to the shared storage for storage therein in response to a first IHS failure, wherein the first IHS failure is an application failure;
• the second HACM in the second IHS being configured to detect first IHS failure and to access most recent first IHS state information from the shared storage in response to the first IHS failure, wherein the first IHS failure is an application failure;
• the second HACM being configured to derive from the most recent first IHS state information a process trace for a failed process executing in the first IHS at the time of first IHS failure, wherein the first IHS failure is an application failure;
• the second HACM being further configured to determine one of the first and second applications in the first IHS as being a closest fit match to the process trace, wherein the determining includes analyzing, by the second HACM, a hierarchy of child processes and parent processes in the process trace for the failed process to determine one of the first and second applications in the first IHS as being a closest fit match to the process trace, thus identifying a particular failed application in the first IHS, as the first IHS failure, wherein the first IHS failure is an application failure.

A computer program product, comprising:

• a computer readable storage medium for use on a first IHS including a first processor, a first memory, a first high availability cluster manager (HACM), and first and second applications in the first memory, and on a second IHS including a second processor, a second memory, and a second high availability cluster manager (HACM), wherein the first and second IHSs couple by a network bus, wherein the first and second IHSs couple to a shared storage, the shared storage communicating with the first and second HACMs;
• first instructions that cause the first IHS to dump first IHS state information to the shared storage in response to a first IHS failure, wherein the first IHS failure is an application failure;
• second instructions wherein the second HACM detects first IHS failure, wherein the first IHS failure is an application failure, and accesses most recent first IHS state information from the shared storage in response to the first IHS failure, wherein the first IHS failure is an application failure;
• third instructions wherein the second HACM derives from the most recent first IHS state information a process trace for a failed process executing in the first IHS at the time of first IHS failure, wherein the first IHS failure is an application failure;
• and
• fourth instructions wherein the second HACM determines one of the first and second applications in the first IHS as being a closest fit match to the process trace, wherein the fourth instructions include instructions that analyze, by the second HACM, a hierarchy of child processes and parent processes in the process trace for the failed process to determine one of the first and second applications in the first IHS as being a closest fit match to the process trace, thus identifying a particular failed application in the first IHS as the first IHS failure, wherein the first IHS failure is an application failure;
• wherein the first, second, third and fourth instructions are stored on the computer readable storage medium.