Synchronous replication error detection and handling

摘要

Some embodiments are directed to a method, corresponding system, and corresponding apparatus that may store data and may monitor, detect, and handle one or more warning or error indications within one or more synchronized replication volumes. Some embodiments may provide first and second storage pools of storage devices with respective volumes. In some embodiments, the first and second storage pools may not share the same controller. Some embodiments may synchronize the first and second storage pools by a storage device of the first storage pool. Some embodiments may monitor for failures, including but not limited to warnings, soft errors, and/or hard errors, at a storage device of the first storage pool. In some embodiments, the one or more failures may be invisible or inaccessible to a user. Prior to an out of sync event or failure, some embodiments may automatically replace the second volume with the first volume.

主权项

1. A computer-implemented method for storing data, the method comprising:
providing a first pool of storage having a first plurality of storage devices and an associated first volume; providing a second pool of storage having a second plurality of storage devices and an associated second volume, data of the first pool of storage corresponding to data of the second pool of storage, the first pool of storage and the second pool of storage not sharing the same controller; synchronizing, by a storage device of the first plurality of storage devices, the first pool of storage with the second pool of storage; monitoring, at a storage device of the first plurality of storage devices, each given storage device of the first plurality of storage devices for one or more warning indications including one or more soft errors; and based on results of the monitoring, automatically taking corrective action, prior to a loss of synchronization event and prior to a failure of the first volume, wherein the synchronizing includes performing one or more bi-directional acknowledgements between the first pool of storage and the second pool of storage; and wherein automatically taking corrective action includes replacing the second volume with the first volume, prior to the loss of synchronization event and prior to the failure of the first volume.