GCP Database Engineer flashcards
GCP Database Engineer · Build recall with a focused review session.
32 cards. Read the term and the definition here. The deck at the top of this page is optional.
- Cloud SQL HA failover
- A regional Cloud SQL instance keeps a standby in a second zone sharing the same regional disk. On failure the standby is promoted automatically; clients reconnect using the same connection name.
- Cloud SQL read replica
- An asynchronous copy of a Cloud SQL primary used to offload reads. Replicas can be promoted to primary and may lag behind the primary under write load.
- Spanner external consistency
- Spanner assigns commit timestamps using TrueTime, so transactions appear globally ordered. Readers always see the latest committed data regardless of which replica they contact.
- Spanner multi-region SLA
- A multi-region Spanner configuration replicates across regions and offers a 99.999% availability SLA, tolerating the loss of an entire region without data loss.
- Bigtable replication routing
- Bigtable clusters replicate asynchronously across zones or regions. App profiles with single-cluster routing give strong reads; multi-cluster routing gives eventual consistency and higher availability.
- AlloyDB architecture
- AlloyDB separates compute from a regional storage layer that handles logging and caching. The primary and read-pool replicas share storage, enabling fast failover and read scaling.
- Firestore multi-region mode
- A Firestore database location choice that replicates across regions for 99.999% availability and strong consistency. Regional locations cost less and reduce latency for nearby users.
- Memorystore standard tier
- A Redis tier with a replica in a second zone and automatic failover. The basic tier has no replica and provides no automatic failover guarantee.
- Datastream
- A serverless change data capture service that streams inserts, updates, and deletes from Cloud SQL, AlloyDB, or Spanner into BigQuery or Cloud Storage.
- BigQuery federated query
- EXTERNAL_QUERY lets BigQuery read live tables in Cloud SQL or Spanner without copying data. It is read-only and limited by the source database's latency and capacity.
- BigQuery Omni
- A BigQuery capability that runs queries on data stored in AWS S3 or Azure Blob Storage without moving it. Results can be combined with BigQuery-native tables.
- Change data capture pattern
- Streaming database changes to an analytical store instead of dual-writing. The OLTP database stays authoritative, and consumers apply changes idempotently to avoid inconsistency.
- Spanner change streams
- A Spanner feature that records data changes with timestamps. Reading the stream requires the Data Change Records API, commonly through Dataflow, not a native BigQuery sink.
- VPC Service Controls perimeter
- A security boundary that restricts data movement between projects and services. It prevents managed database data from leaving the perimeter unless explicitly allowed.
- Private Service Connect
- A private endpoint mechanism that lets clients reach managed Google services, including databases, over internal IP addresses without public exposure or internet traversal.
- Database Migration Service
- A managed service for homogeneous and heterogeneous migrations into Cloud SQL, AlloyDB, and Spanner. It uses an initial dump plus continuous replication before a planned cutover.
- Homogeneous migration
- A migration that keeps the same database engine, such as MySQL to Cloud SQL MySQL. Schema and application code usually need little or no change.
- Heterogeneous migration
- A migration that changes the database engine, such as Oracle to PostgreSQL. It requires schema conversion, code changes, and validation of behavior differences.
- HarbourBridge
- A tool that migrates MySQL or PostgreSQL schemas and data into Spanner. It does not convert stored procedures, triggers, or other engine-specific logic.
- Migration validation
- Comparing row counts, checksums, and query results between source and target before cutover. Replication lag and error metrics from the migration tool guide readiness.
- BigQuery Data Transfer Service
- A scheduler for recurring loads from SaaS sources and Cloud Storage into BigQuery. It is not a one-time migration tool for transactional databases.
- Database Migration Program
- Google Cloud tooling and guidance for moving Oracle and SQL Server workloads to PostgreSQL. It covers schema conversion, code assessment, and data movement.
- Storage Transfer Service
- A service for moving large datasets from on-premises or other clouds into Cloud Storage. Loaded files can then be imported into BigQuery or other targets.
- Cloud SQL storage auto-increase
- A setting that grows a Cloud SQL disk when free space runs low, up to a configured limit. It prevents write outages caused by a full disk.
- Spanner processing units
- The unit of Spanner capacity. One node equals 1000 processing units; regional instances need at least one node, and multi-region instances need at least three.
- Bigtable instance and clusters
- A Bigtable instance contains one or more clusters, each in a single zone. Multi-cluster routing requires at least two clusters for availability and failover.
- AlloyDB read pool
- A set of read replicas added to an AlloyDB cluster to scale reads. The pool is configured separately from the primary instance and shares the storage layer.
- Cloud SQL point-in-time recovery
- Restoring a Cloud SQL instance to a specific moment using automated backups plus transaction logs. It requires binary logging to be enabled on the instance.
- Private services access
- A VPC peering connection between your network and Google's network that enables private IP access to Cloud SQL and Memorystore. It requires an allocated IP range.
- Spanner backup schedule
- A Spanner feature that creates backups automatically on a defined schedule and retention period, supporting restore to a new instance or the same instance.
- BigQuery table clone
- A lightweight copy of a BigQuery table that shares storage with the source. It is cheap to create and useful for testing before promoting changes.
- Terraform database resources
- Infrastructure-as-code resources such as google_sql_database_instance and google_spanner_instance that create and manage databases reproducibly across environments.