etcd vs Google Cloud Bigtable

etcd is an open-source, strongly consistent distributed key-value store for data used by distributed systems and clusters of machines. It arranges values in hierarchical directories and lets clients watch keys or directories for changes. Clients can read and write values with standard HTTP tools such as curl, and keys may have optional time-to-live expiration. The project uses the Raft protocol and handles leader elections during network partitions, including when a leader machine fails. etcd is a core Kubernetes component and its primary datastore. The project provides pre-built binaries, source builds, and container images. Linux on AMD64 and ARM64 is tier 1; Darwin and Windows on either architecture are tier 3, and other platforms are unsupported. TLS and client-certificate authentication are supported for client and peer connections, but neither is enabled by default. The software is self-hosted and licensed under Apache 2.0, with a listed price of free. The integrations listing includes Kubernetes, CoreDNS, and language client libraries, while warning that third-party tools and libraries there are not tested or maintained by the etcd team.

Google Cloud Bigtable is a managed, low-latency NoSQL database for machine learning, operational analytics, and user-facing applications. It stores structured, semi-structured, and unstructured data in key-value and wide-column formats. SQL querying includes an API with more than 200 functions and support for aggregations, windowing, geospatial and JSON processing, and vector search. Compute scales horizontally, and instances can span up to eight regions with automatic replication between clusters. Google says multi-region instances offer 99.999% availability and protection against a regional failure. Integrations listed include Apache Spark, Hadoop, GKE, Dataflow, Dataproc, Vertex AI Vector Search, BigQuery, and LangChain. Bigtable provides Apache HBase and Cassandra APIs, plus migration tools for HBase, Cassandra, DynamoDB, and Aerospike. Security controls include IAM, customer-managed encryption keys, audit logging, and table-, column-, or row-level access control. Deployment is cloud-based. Google describes its capacity-based low-latency compute as more favorable for larger workloads than small applications. A 10-day trial is available subject to eligibility requirements.