| CPC G06F 11/2069 (2013.01) [G06F 11/0772 (2013.01); G06F 11/1453 (2013.01); G06F 11/2023 (2013.01); G06F 11/3034 (2013.01)] | 20 Claims |

|
1. A computer-implemented method for a non-disruptive automatic unplanned failover performed by one or more processors of a multi-site distributed storage system with a primary storage site having a first cluster and a secondary storage site having a second cluster, the method comprising:
initiating an out of sync (OOS) state for a relationship between the first and second clusters when the secondary storage site fails to receive heartbeat information from the primary storage site during a time period with the heartbeat information indicating an operational condition for the primary storage site;
initiating, with a mediator, an automatic unplanned failover based on the OOS state; storing an indication of the mediator for initiating the automatic unplanned failover; and
completing operations of the automatic unplanned failover with restorability based on storing the indication of the mediator initiating the automatic unplanned failover, which enables the automatic unplanned failover to restart and have a successful outcome for the automatic unplanned failover even when multiple failures occur in the multi-site distributed storage system.
|