SYSTEM Cited by 1 source

Meta Hive (data warehouse)

Meta Hive is Meta's internal long-term / cold-storage data warehouse, derived from the original Facebook Hive paper (2010, "Hive — A Warehousing Solution Over a Map-Reduce Framework"). Cited by name in the 2024-12-02 cryptographic monitoring post as Meta's canonical cold-tier for telemetry that outgrows warm storage (Scuba).

The open-source descendant is Apache Hive; Meta's internal Hive has diverged but the paper is the shared origin. This wiki page is a stub for Meta's internal variant — use systems/apache-hive for the general-purpose Apache Hive concept.

Role for this wiki

• Cold tier in Meta's telemetry storage pattern. Scribe → Scuba (warm, real-time) + Hive (cold, long-retention) is the canonical Meta two-tier telemetry stack disclosed in the 2024-12-02 post.
• Long retention window. The post ties Meta's ability to "monitor trends over time" and do "predictive modeling and analysis" on cryptographic usage to the long retention window of the Hive tier.
• Primary warehouse for Meta's internal analytics. Cited as such throughout Meta's engineering corpus when long-term storage of logged events is referenced.

Seen in

• sources/2024-12-02-meta-built-large-scale-cryptographic-monitoring — Hive is the cold-tier destination for FBCrypto's flushed aggregated counts; enables long-retention trend analysis of cryptographic-primitive usage.

Related

• systems/apache-hive — open-source descendant of the same 2010 paper.
• systems/scuba-meta — warm sibling tier.
• systems/scribe-meta — upstream ingestion.
• systems/fbcrypto — one producer.