Failure chain detection and recovery in a group of cooperating systems

A method, comprising:

• a group management service (GMS) module one of a plurality of computer nodes cooperating as a group according to a group management services protocol detecting that one of the plurality of nodes failed, wherein the failed node was performing recovery operations for a previously failed node when it failed;
• the GMS module selecting one of the plurality of computer nodes to perform recovery operations for the failed node;
• the GMS module updating group state data, accessible to each of the plurality of nodes, to indicate the selected node as selected to perform recovery operations for the failed node, wherein the group state data comprises information regarding the recovery state of each of the plurality of nodes;
• and
• a GMS module on the selected node requesting a client process on the selected node to perform the recovery operations for the failed node.

A computer-accessible memory medium, comprising program instructions computer-executable to implement:

• a group management service (GMS) module one of a plurality of computer nodes cooperating as a group according to a group management services protocol detecting that one of the plurality of nodes failed, wherein the failed node was performing recovery operations for a previously failed node when it failed;
• the GMS module selecting one of the plurality of computer nodes to perform recovery operations for the failed node;
• the GMS module updating group state data, accessible to each of the plurality of nodes, to indicate the selected node as selected to perform recovery operations for the failed node, wherein the group state data comprises information regarding the recovery state of each of the plurality of nodes;
• and
• a GMS module on the selected node requesting a client process on the selected node to perform the recovery operations for the failed node.