| Popis: |
In recent years, NoSQL databases have emerged as an answer to Internet scale datasets. However, these solutions do not support out of the box the necessary backup, restore and disaster recovery capabilities required by many big data scenarios. In this paper we address this gap: we present a novel disaster recovery approach designed for big data NoSQL workloads. This approach can support low recovery point and time objectives at low costs. Our approach is highly scalable and elastic, based on open source components and suitable for any document-based NoSQL solution, which makes it highly attractive for diverse big data workloads. |
