We are trialing the use of BigTable and are doing a capacity planning exercise. We think one node will satisfy our needs in the early stages of our implementation and we add more as necessary later. My only question is, do we require more than one node in order to provide high availability? (My lack of understanding of what constitutes a "node" is obvious here, the answer to this question will help me to understand)
1 Answers
To provide high availability of Cloud Bigtable instances, you need more Clusters, not nodes. A cluster represents the Bigtable service in a specific location and each cluster is located in a single zone. To attain high availability of your instance you need to create more cluster. You can also visit Google recommendation to know more about creating Cloud Bigtable High Availability.
Note: Nodes in the cluster of Cloud Bigtable instances are the compute resources that Bigtable uses to manage your data. If you increase the number of nodes, you are increasing the performance of your cluster to support the larger workload.
Related