Recovery mechanisms across storage nodes that reduce the impact on host input and output operations

摘要

Provided are a method, a system, and a computer program product in which a storage controller determines one or more resources that are impacted by an error. A cleanup of tasks associated with the one or more resources that are impacted by the error is performed, to recover from the error, wherein host input/output (I/O) operations continue to be processed, and wherein tasks associated with other resources continue to execute.

主权项

1. A method, comprising:
determining, by a storage controller, one or more resources that are impacted by an error; and performing a cleanup of tasks associated with the one or more resources that are impacted by the error, to recover from the error, wherein host input/output (I/O) operations continue to be processed, and wherein tasks associated with other resources continue to execute, the method further comprising:
determining whether an impacted resource is in an I/O path;in response to determining that the impacted resource is in the I/O path, determining whether the impacted resource is currently executing I/O;in response to determining that the impacted resource is currently executing I/O, performing an I/O recovery sequence; andin response to determining that the impacted resource is currently not executing I/O, determining whether operations are being performed in a non I/O recovery sequence for the impacted resource.