Block storage
Local SSD storage for virtual machine instances
High-performance, ephemeral block storage.
功能
Fastest IOPS and high-speed throughput
Choose Local SSDs when you need Google's highest-speed ephemeral storage. Reach IOPS of 3,200,000 / 1,600,000 (read / write) and throughput up to 12,480 MiB per second / 6,240MiB per second on general purpose instances for 12 TB instances. On the newest storage-optimized instances, reach an industry-leading IOPS of 6,000,000 read and write and throughput up to 36,000 MiB per second / 30,000 MiB per second (read / write) on 36 TB instances.
Scale as you need it
Attach up to 32 Local SSD disks for 12 TB of total Local SSD storage space per general purpose instance, you can format and mount multiple Local SSD partitions into a single logical volume.
Customizable VM shapes
Depending on your workload, you may require specific memory-to-storage configuration to achieve the best performance at the right price. With Local SSDs, attach custom-sized disks to your VMs, allowing you to tailor your storage to your use case, needs, and budget.
工作方式
Local SSD in a minute
Wanting to find a tool that gives you extra storage for your VM instances? In this episode of Cloud Bytes, Jen Person explains what a Local SSD is and the different use cases for this tool. Watch to learn if this ephemeral storage option fits best with your developer projects.
Local SSD in a minute
Wanting to find a tool that gives you extra storage for your VM instances? In this episode of Cloud Bytes, Jen Person explains what a Local SSD is and the different use cases for this tool. Watch to learn if this ephemeral storage option fits best with your developer projects.
常见用途
Caches or storage for transient, low value data
Using Local SSDs for caching allows you to store data in a lower-latency drive so responses to requests for that data can occur fast.
学习资源
Using Local SSDs for caching allows you to store data in a lower-latency drive so responses to requests for that data can occur fast.
Scratch processing space for high performance computing or data analytics
Local SSDs are designed for temporary storage use cases such as scratch processing space. Because Local SSDs are located on the physical machine where your VM is running, they allow for very fast temporary storage space for HPC or data analytics.
学习资源
Local SSDs are designed for temporary storage use cases such as scratch processing space. Because Local SSDs are located on the physical machine where your VM is running, they allow for very fast temporary storage space for HPC or data analytics.
Temporary data storage for Microsoft SQL Server
Local SSDs can be a high-performance option for Microsoft SQL Server's tempdb system database and Windows pagefile. Local SSDs are physically connected to the server that hosts the VM instance, which can lead to better performance, lower latency, and higher input/output operations per second (IOPS) compared to other block storage options. However, local SSDs are removed when the instance is shut down or reset using the API.
学习资源
Local SSDs can be a high-performance option for Microsoft SQL Server's tempdb system database and Windows pagefile. Local SSDs are physically connected to the server that hosts the VM instance, which can lead to better performance, lower latency, and higher input/output operations per second (IOPS) compared to other block storage options. However, local SSDs are removed when the instance is shut down or reset using the API.
价格
| How Local SSD pricing works | Pricing for Local SSDs is based on usage, and is typically billed per GB per month. |
|---|---|
| Product | Pricing |
Local SSD |
How Local SSD pricing works
Pricing for Local SSDs is based on usage, and is typically billed per GB per month.
Get started with Local SSDs
Learn more about options and pricing
New customers get $300 in free credits on signup
Learn how to optimize Local SSD performance
How to choose a disk interface
Understand how to choose a valid number of disks
业务用例
Top companies are using Local SSD
Mixpanel: Scalable, high-performance product analytics
Joe Xavier, VP of Engineering, Mixpanel
"We saw the pace of innovation as faster at Google, which opens up possibilities for us to do things we otherwise couldn’t do. With Google Cloud doing a lot of the heavy lifting, we can concentrate on making our analytics products the best they can be."
Learn more
