CnosDB vs Grafana Mimir

CnosDB is an open-source distributed time-series database for workloads such as IoT, industrial internet, connected vehicles, and IT operations. It supports standard SQL, time-based aggregation, and a DataFrame API. Data can be written without a predefined schema, including out of order, and historical data can be replenished. Ingestion methods include SQL, Line Protocol, OpenTSDB, Elasticsearch bulk, Loki, NDJSON, and OpenTelemetry. Deployment options include standalone, distributed, and storage-compute separated modes, on public, private, or hybrid clouds. The documentation describes horizontal scaling for storage and stateless compute nodes. Integrations include Grafana, Prometheus, Telegraf, JDBC, and a Python SDK; the open-source Python connector provides a PEP 249 interface for SQLAlchemy and pandas. Import and export formats include PARQUET, CSV, and JSON with AWS S3, Google Cloud Storage, and Azure Blob Storage. CnosDB documents retention policies, multi-tenancy, and role-based permissions, as well as optional TLS certificate and private-key settings. Community Edition is free under AGPL-3.0.

Grafana Mimir has no write-up on Specifiction yet; its figures are above.