A storage pool is a set of physical storage devices in a protection domain. A volume is distributed over all devices residing in the same storage pool. You can define a magnetic storage pool (for HDDs) or a high performance storage pool (for SSDs). Storage pools support medium or fine granularity data layouts and allow enabling or disabling zero padding.
Storage pools allow the managing of different storage tiers in
PowerFlex. Each storage device belongs to one (and only one) storage pool. The figure shows two storage pools.
When a volume is configured over the virtualization layer, it is distributed over all devices residing in the same storage pool. Each volume block has two copies on two different fault units. This allows the system to maintain data availability following a single-point failure. Two network failures render all the domain in a state where I/O errors are possible on any storage pool.
You must assign a media type setting to each storage pool. Supported types are: HDD, SSD, and Transitional (allows for migration flows).
If all SDSs in a protection domain have two physical drives associated with them, such as one hard drive, and the other SSD, you should define two storage pools:
Magnetic storage pool
Consists of all HDDs in the protection domain
High performance storage pool
Consists of all SSDs used for storage purposes in the protection domain
NOTE:Mixing different types of SSDs is not recommended. Creating a separate storage pool for each type is the recommended best practice, for example: SAS SSD, SATA SSD, NVMe SAS SSD.
NOTE:PowerFlex might not perform optimally if there are large differences between the sizes of the fault units in the same storage pool. For example, if one device has a much larger capacity than the rest of the devices, performance may be affected. After adding devices, you can define how much of the device capacity is available to
PowerFlex by using the SCLI
modify_sds_device_capacity command.
Storage pools support the following data layouts for HDD or SSD media:
Medium granularity (MG): Space allocation occurs at 1 MB units.
Supports HDD and SSD media
Includes persistent checksum for data integrity
Fine granularity (FG):
Requires SSD media and NVDIMM for acceleration
Space allocation occurs at 4 KB units
Includes persistent checksum for data integrity
Supports data compression which reduces the size of data that is stored on the disk
Supports thin-provisioned, zero-padded volumes
NOTE:FG and MG storage pools can both exist in a single SDS. You can also migrate volumes across the two layouts.
Each storage pool can work in one of the following modes:
Zero-padding enabled
Ensures that every read from an area previously not written to returns zeros. Some applications might depend on this behavior. Furthermore, zero padding ensures that reading from a volume will not return information that was previously deleted from the volume.
This behavior incurs some performance overhead on the first write to every area of the volume since the area must be filled with zeros first.
FG is always zero padded.
Zero-padding disabled (default only for MG)
A read from an area previously not written to will return unknown content. This content might change on subsequent reads.
Some applications assume that when reading from areas not written to before, the storage will return zeros or consistent data. If you plan to use such applications, then zero padding must be enabled.
You can add storage pools during installation. In addition, you can modify storage pools post-installation with most of the management clients.
NOTE:The zero padding policy cannot be changed after the addition of the first device to a specific storage pool.
Data is not available for the Topic
Please provide ratings (1-5 stars).
Please provide ratings (1-5 stars).
Please provide ratings (1-5 stars).
Please select whether the article was helpful or not.
Comments cannot contain these special characters: <>()\