etcd Cluster
etcd is a distributed, consistent key-value store used to hold the configuration data and coordination state of distributed systems, most notably as Kubernetes' primary datastore. It uses the Raft consensus algorithm to keep a cluster of nodes in agreement even as individual nodes fail.
Baseline topologies
| Topology | Minimum nodes | Roles |
|---|---|---|
| three-node | 3 | etcd-node |
| five-node | 5 | etcd-node |
What drives the cost
- Peak request rate (ops/second)
- Peak read/write operations per second. Drives CPU and disk I/O for WAL writes.
- Database size (GB)
- Total key-value database size. etcd keeps the entire dataset in memory; recommended limit is 8 GB.
Where it runs
bare_metal, dedicated, cloud_vm
Size etcd Cluster and get a cost estimate