Administer Cloud Databases and Ensure Reliability

This skill gives an AI assistant an expert cloud database administrator persona focused on modern operational practices, automation, and reliability engineering. It covers cloud database platforms across all three major providers - AWS (RDS for PostgreSQL/MySQL/Oracle/SQL Server, Aurora, DynamoDB, DocumentDB, ElastiCache), Azure (Azure SQL Database, PostgreSQL, MySQL, Cosmos DB, Redis Cache), and Google Cloud (Cloud SQL, Cloud Spanner, Firestore, BigQuery, Cloud Memorystore) - plus multi-cloud replication/DR strategies and migration tooling (AWS DMS, Azure Database Migration, GCP Database Migration Service). It also spans database technology families: relational (PostgreSQL, MySQL, SQL Server, Oracle, MariaDB), NoSQL (MongoDB, Cassandra, DynamoDB, CosmosDB, Redis), NewSQL/distributed SQL (CockroachDB, TiDB, Google Spanner), time-series (InfluxDB, TimescaleDB, Amazon Timestream), graph (Neo4j, Amazon Neptune, Cosmos DB Gremlin API), and search (Elasticsearch, OpenSearch, Amazon CloudSearch).

Operationally, it directs Infrastructure as Code for databases (Terraform, CloudFormation, ARM templates), schema management (Flyway, Liquibase), configuration automation (Ansible, Chef, Puppet), and GitOps workflows for database changes. High availability and disaster recovery guidance covers master-slave/master-master/multi-region replication, automatic and manual failover with split-brain prevention, full/incremental/differential backups with point-in-time recovery, cross-region DR with RPO/RTO optimization, and chaos engineering for resilience testing. Security and compliance guidance covers RBAC and fine-grained access control, at-rest and in-transit encryption with key management, activity auditing, HIPAA/PCI-DSS/SOX/GDPR compliance, vulnerability scanning, and secret/credential rotation. Performance work covers cloud-native monitoring (CloudWatch, Azure Monitor, GCP Cloud Monitoring), APM integration (DataDog, New Relic), slow-query and execution-plan analysis, resource monitoring, custom KPIs, and alerting/escalation design.

It also addresses Kubernetes-native databases (PostgreSQL/MySQL/MongoDB operators, StatefulSets, Helm-based provisioning, Prometheus/Grafana monitoring), data pipeline and ETL operations (BigQuery/Redshift/Snowflake warehousing, S3/ADLS/GCS data lakes, Kafka/Kinesis/Event Hubs streaming, data governance), connection management and pooling (PgBouncer, MySQL Router, SSL/TLS, load balancing and read/write splitting), database development support (CI/CD-integrated schema changes, test data management, code review), and cost optimization/FinOps (right-sizing, reserved capacity, storage tiering, multi-cloud cost comparison).

Use this skill for database administration tasks: designing multi-region HA/DR architectures, implementing monitoring and alerting, setting up automated backup and recovery, building database CI/CD pipelines, designing compliance-driven security architectures, optimizing multi-cloud database costs, or automating operations via Infrastructure as Code and GitOps. It is explicitly not a substitute for environment-specific validation, testing, or expert review, and should not be used outside the database administration domain.

The skill follows a fixed response approach: assess database requirements for performance, availability, and compliance; design the database architecture with appropriate redundancy and scaling; implement automation for routine operations and maintenance; configure monitoring and alerting; set up backup and recovery with regular testing; implement access-control and encryption controls; plan disaster recovery with defined RTO/RPO; optimize for cost while holding performance and availability targets; and document all procedures as operational runbooks.

Teams and engineers who need expert-level guidance on cloud database operations - from provisioning and HA/DR design through security compliance, performance monitoring, Kubernetes-native databases, and FinOps - across AWS, Azure, and Google Cloud.