Intrigued by the art of cloud architecture? Discover how to design, develop, and manage robust, secure, scalable, and dynamic solutions on Google Cloud as you prepare for the Professional Cloud Architect exam!
Gauge your current knowledge
Gauge your current knowledge
Google Cloud provides block storage options like Persistent Disks and Hyperdisk, which act like virtual hard drives for your virtual machines. These disks can be scaled up in size to handle more data without needing to delete your virtual machines. Performance for storage is measured by IOPS (the number of read/write operations per second) and throughput (the amount of data transferred per second). For many disk types, these performance limits increase automatically as you increase the disk's capacity, so larger disks generally provide higher speed.
Hyperdisk offers more flexibility by letting you set the IOPS and throughput independently from the disk's total storage size. You can manage this using Hyperdisk Storage Pools, which group storage resources for multiple virtual machines in a zone to ensure capacity is available and efficiently used. This is especially helpful for workloads that need precise performance tuning.
A key decision is choosing between zonal and regional storage, which involves a trade-off between cost and data durability. Regional Persistent Disks keep your data safe by synchronously copying it across two zones within a region, protecting against a failure in one data center. This is more expensive than a zonal disk, which only exists in one zone and is less resilient. To prepare for disasters, you can use reservations to lock in guaranteed capacity in a specific zone and use regular snapshots to create incremental backups of your disks, allowing for quick recovery.
Services like Cloud SQL, Spanner, and Bigtable are managed databases, meaning Google handles much of the underlying infrastructure. These services include features for automatic scaling and monitoring. To handle growing data, you can enable storage auto-resize, which automatically increases capacity when the database is nearly full. For demanding performance needs, Google Cloud Hyperdisk lets you provision IOPS and throughput separately to match your specific workload, helping keep the system stable during traffic spikes.
Different storage options are available for these managed systems. Hyperdisk Balanced fits most general-purpose workloads, while Hyperdisk Extreme is designed for databases needing the highest possible IOPS. Persistent Disk offers reliable, scalable block storage.
High availability is achieved through data replication. For example, Spanner can use multi-region configurations to copy data across different geographic areas, protecting against regional outages. An HA-configured instance can automatically switch to a standby database in another zone if the primary one fails. Managing provisioned throughput is also crucial; you reserve the capacity needed to handle a certain number of requests per second to prevent service interruptions and errors from exceeding limits.
Cost efficiency is managed by choosing the right storage classes based on how often data is accessed. You can use Object Lifecycle Management to create rules that automatically move older, less-used data to cheaper storage tiers over time, ensuring your budget is spent wisely.
Cloud Storage offers different storage classes to balance cost and how quickly you can access your data. Standard storage is for frequently accessed data. Classes like Nearline, Coldline, and Archive are for data accessed less often, with Archive being the cheapest for long-term storage. Choosing the right class means you pay only for the performance level your application requires.
You also choose a location type for your data buckets based on residency rules and availability needs. Regional storage keeps data in one geographic area for low latency. Dual-region stores data in two separate regions, and multi-region spreads it across at least two locations for the highest protection against large-scale disasters.
Object Lifecycle Management automates moving data between these storage classes to save money. You set rules based on the age of the data or the last time it was accessed. For example, you can have logs automatically move to a cheaper class after 30 days. This automation reduces manual work and optimizes costs.
When planning storage, you must weigh the trade-offs. Colder storage classes have lower monthly costs but higher fees for retrieving the data. Understanding your data access patterns lets you optimize this balance. Data durability is ensured through replication; Google Cloud keeps copies of your data in different physical zones. Using multi-region buckets provides the strongest protection, keeping your data available even if an entire region experiences an outage.
StudyCerts
StudyCerts