Alibabacloud Odps Information Schemav0.0.1

FinOps and data platform teams query QUOTA_USAGE, TASKS_HISTORY, and cost_cpu/cost_mem columns to understand which projects, jobs, or users drive compute consumption. They convert cost_cpu/100/3600 into CU-hours and trend usage over time to catch runaway pipelines before the monthly bill surprises them.

Security and compliance teams use TABLE_PRIVILEGES, COLUMN_PRIVILEGES, USERS, and USER_ROLES to answer questions like 'who has access to table X' and to audit expired or overly broad grants. The static snapshot nature of privilege views suits periodic attestation reports rather than real-time monitoring.

Data engineers investigate failed or slow jobs by filtering TASKS_HISTORY with a ds partition filter and status = 'Failed', then examining start/end times via DATEDIFF to find runtime hotspots. TUNNELS_HISTORY complements this for upload/download tunnel audit trails.

Platform administrators identify unused or oversized tables by joining TABLES, COLUMNS, and PARTITIONS views, using COALESCE(last_access_time, last_modified_time) for partitioned tables. This helps reclaim storage and prioritise retention policies across dozens of projects in a metadata center.

Analysts build cross-project inventories of catalogs, schemas, volumes, and foreign servers that only tenant-level Information Schema exposes. They use the mandatory 3-field join key (table_catalog, table_schema, table_name) to correlate metadata correctly across a multi-catalog environment.

Customers pay for the compute resources (CU-hours, derived from cost_cpu and cost_mem) and storage they consume, so fine-grained Information Schema metadata directly maps to invoiced cost. Usage insights from TASKS_HISTORY and QUOTA_USAGE feed capacity planning and cost optimisation conversations.

Alibaba Cloud packages metadata auditing, permission attestation, and lineage tooling as value-added governance services for enterprise accounts, monetised via support tiers and premium platform editions. Information Schema acts as the underlying data source these services query.

Service providers run the MaxCompute platform and metadata diagnostics on behalf of clients, monitoring failed tasks, tunnel activity, and permission drift for a recurring fee. The Information Schema queries power dashboards and alerts that justify the managed-service contract.