Netapp, Inc.

Yi Zhang

Emalayan Vairavanathan

Dheeraj Raghavender Sangamkar

Song Guen Yoon

Tymoteusz Altman

Techniques are described for reserving space on a destination node or volume for increasing the likelihood of a successful data transfer in a distributed storage environment. A reservation may be retried at one or more destinations if the reservation fails at a first destination. In some embodiments, the data-transfer process can be paused or terminated prior to data being transferred to one or more destinations if a reservation fails. Reserving space on a destination node or volume can increase the likelihood of a successful data transfer, which can increase the likelihood of efficient resources usage in a storage system.

Space reservation for distributed storage systems

To ensure that there is an elected manager among storage nodes of an erasure coding group ('ECG'), an ECG manager ('ECGM') election process is periodically performed among available storage nodes that are configured with the software to perform the services of an ECGM. When a storage node is activated, an ECGM process of the storage node begins executing and is assigned a process identifier (&#x201c;PID&#x201d;). A storage node can utilize a service query framework to identify other available storage nodes and retrieve their ECGM PIDs. The storage node then selects a PID according to a criterion and elects the storage node corresponding to the selected PID to be the acting ECGM. This process is performed periodically, so even if the acting ECGM storage node fails, a new ECGM is eventually selected from the available storage nodes.

Manager election for erasure coding groups

To efficiently recover from a multiple storage node failure, a storage node concurrently restores data fragments to the multiple failed storage nodes, as opposed to restoring each node individually. In the VCS based storage technique, storage nodes are restored as part of an ECG repair process. For each ECG being repaired, a storage node performing the restoration process reads data fragments from active nodes in the ECG and generates new data fragments to replace any lost data fragments. The node then stores one of the new data fragments across each of the failed storage nodes. By concurrently restoring data fragments to each failed storage node, the data fragments needed to repair each ECG are only read once, thereby preserving disk operations and network bandwidth.

Growing community of inventors

Vancouver, Canada

Yi Zhang