Can Your Witness Do This?
The majority of edge sites require high availability as a priority. When you don’t have a team on site to respond to unplanned downtime quickly, you’re leaving yourself open and disaster could strike at any time. In these cases, it is sensible to design your storage infrastructure with sufficient redundancy and eliminate any single point of failure. This will create some protection from the threat of unplanned downtime. But what happens if you lose an entire node in a cluster? If you are to truly mitigate against these types of critical failures and the subsequent downtime, while maintaining a lightweight, simple IT infrastructure, then you need the SvSAN witness.
What does the ‘witness’ do?
The witness is a passive element of SvSAN. Its sole purpose is to maintain the mirror state of a cluster, should one of the nodes go offline unexpectedly. The witness can sit remotely from the cluster nodes and its low hardware requirements mean it can run on something as small as a Raspberry Pi. No need for a third node in every cluster – a single SvSAN witness can maintain the mirror state of hundreds or even thousands of clusters at the same time.
The SvSAN witness simplifies clustered environments. The technology seamlessly avoids ‘split-brain’ (which could lead to data corruption) by monitoring each cluster for health status and if one of the nodes goes down unexpectedly, the witness makes sure the surviving node assumes the role as ‘leader’. If a leader cannot be determined, only then is the cluster taken offline, as to avoid data corruption. This means that if one of the nodes goes down then your applications will (in most cases) keep running as usual, while you get the failed node back online, causing minimal disruption. Once the failed node is repaired, the data is automatically synchronized and full protected mode resumes.
The witness is incredibly resilient and can withstand a variety of failure scenarios, a full analysis of which can be found on the SvSAN Witness white paper. The white paper provides a complete overview of the feature and also includes system requirements and performance figures.For more information on how you can increase the resilience of your storage cluster, get in touch with the team at [email protected] and ask about the SvSAN witness.
