“Thin Provision”, also known as “over-allocating”, is an important emerging storage technology that automatically optimizes configuration. This article explores automated simplified configuration, explaining how it works, limitations of use, and recommendations for use.
If the storage space used by the application is full, it will crash. To avoid potential application failures, storage administrators often allocate more storage capacity to applications than is actually needed. This approach leaves room for future growth and reduces the risk of application failure. But it requires more physical disk capacity than actual, causing waste.
This automated provisioning software eliminates the need to allocate unused disk space in advance, thereby improving storage utilization. Figure 1 shows how storage administrators traditionally allocate more storage space than applications need - planning ahead for future growth and ensuring applications don't crash because disk space is full. Volume A in Figure 1 only has 100G of physical data, but may be allocated much more storage based on growth projections (500GB, in this example). The unused storage allocated to this volume cannot be used by other applications. In many cases, these 500 GB are never used and are essentially wasted. This situation is called "stranded storage."
In layman's terms, automatic thin provisioning is equivalent to blowing up a balloon in the same room. As long as there is space in the room, you can blow up the balloon. It is a kind of on-demand allocation.
In most cases, automated storage pool configuration will follow the basic principle of "providing storage from a common storage pool according to the needs of the application." To effectively utilize thin provisioning technology, it can be combined with storage virtualization technology. With thin provisioning, storage administrators can allocate logical storage (e.g. 500G) to applications as usual, but only actually occupy the physical capacity when needed. When the storage utilization approaches a predetermined threshold (for example, 90%, 450G), the array automatically allocates space from the virtual storage pool to expand the volume without manual intervention from the storage administrator. Volumes can be overallocated as usual, so the application thinks it has plenty of storage, but no storage is actually wasted. Basically, thin provisioning technology eliminates the waste of allocated but unused space.
Automatic thin provisioning technology also has some limitations. It is currently not recommended to be used in the following situations:
Data deleted from the volume needs to be recycled, which will cause storage The overhead and cost of the controller increases.
File systems that use unused blocks instead of reusing freed blocks (such as NTFS Microsoft files) can cause the volume to grow beyond its maximum allocated size before reusing the storage. This File systems reduce the benefits of thin provisioning.
Applications whose metadata spans the entire volume will reduce the benefits of thin provisioning.
If the application expects the data to be contiguous, and/or optimizes I/O performance around that assumption, using thin provisioning is not a good option.
If the array cannot provide speeds fast enough to provide new live storage, applications will not run successfully. Try to use high-performance storage controllers and well-monitored storage over-provisioning minotor to avoid reducing availability.
With the development of automated optimization configuration technology in the future, applications and file systems may be redesigned and modified to avoid these problems. The economic rationale for thin provisioning is simple: it automates storage allocation, thereby significantly reducing the work of storage administrators, and it can reduce the amount of storage required to serve applications. It also reduces the number of disks required, resulting in a significant reduction in energy consumption.
When setting up a virtualized data center, the main advantage of automated down-provisioning should be carefully considered, which is to improve the overall utilization of storage space. However, users need to be aware of the caveats and carefully check the storage and management requirements of the application to avoid making the wrong choice.
The above is the detailed content of Example analysis of Thin Provision. For more information, please follow other related articles on the PHP Chinese website!