Home > Storage > PowerMax and VMAX > Storage Admin > Dell PowerMax and VMware vSphere Configuration Guide > Thin device data reduction
The PowerMax enables compression and deduplication through the feature called data reduction. Data reduction is enabled at the storage group level and uses both a hardware and software implementation. This data reduction is a combination of compression and deduplication. These two capabilities cannot be separated as they work together to achieve the desired savings, however each capability will be explained separately below.
On the PowerMax, inline compression is available. Compression on the PowerMax is hardware-based, with the ability to use software in the event of a failure. Individual devices cannot be compressed, rather compression is enabled at the storage group level for all devices within the group. Data is compressed as it is written to disk when the device is in a storage group with compression enabled. If compression is enabled on a storage group that already has data, that data is compressed by a background process. If compression is disabled on a storage group, new written data is not compressed but existing data is not immediately decompressed. Only when data is written again is it decompressed.
Note: Pre-allocated, persistent devices cannot be compressed.
Note: Data reduction is not applicable on data that is stored on Storage Class Memory (SCM).
Deduplication works in conjunction with inline compression to further improve efficiency in the use of storage space. It reduces the number of copies of identical tracks that are stored on back-end devices. Depending on the nature of the data, deduplication can provide additional data reduction over and above the reduction that compression provides. Because deduplication works in conjunction with compression on the PowerMax, one feature cannot be enabled without the other. When the PowerMax detects a duplicated track in a storage group, deduplication replaces it with a pointer to the track that already resides on back-end storage.
PowerMax deduplication with VMware is extremely effective at saving space when copies of virtual machines are made. Users can expect near one hundred percent deduplication.
Compression and deduplication on the PowerMax, as previously mentioned, is referred to as Data Reduction. It is complementary to thin provisioning. Not only can one oversubscribe storage initially, but as the data on the back-end can be compressed and deduped reducing the data footprint, it increases the effective capacity of the array.
Take the following example, Figure 102, where 1.3 PB of storage is presented to hosts in the form of thin devices. The actual physical storage behind the thin devices is only 1.0 PB. Now, by adding data reduction, the physical space required is reduced by half, providing a 2:1 ratio. The array, therefore, requires half as many drives to support the same front-end capacity.
Despite the space savings, it is important to remember that if/when data is decompressed, additional space is required. This is considered when new arrays are sized.
By default, when creating a storage group through Unisphere for PowerMax, the Data Reduction check box is enabled as in and Figure 103. Recall that if compression is enabled, deduplication is also enabled on PowerMax. PowerMax simply refers to the two features as “Data Reduction”. There is no separate check box for deduplication as enabling and disabling compression includes deduplication. Note that when using a parent storage group, it is possible to set data reduction differently for each child storage group.
Figure 103 Unisphere for PowerMax - storage group data reduction
If using Solutions Enabler to create a storage group, data reduction will be enabled by default provided the SRP is supplied on the command line. For instance, in Figure 104 storage group data_reduction_sg is created in SRP srp_1. Therefore, compression is enabled.
Figure 104. Solutions Enabler - storage group compression
If the SRP is left out of command, the storage group no_data_reduction_sg will be created without compression as in Figure 105.
Figure 105. Solutions Enabler - storage group no compression
If the user attempt to create the storage group with the compression flag, -compression, while still not providing an SRP, Solutions Enabler will indicate it is not possible since the group is not FAST managed. This is seen in Figure 106.
Figure 106. Solutions Enabler - storage group non-FAST
Unisphere for PowerMax and Solutions Enabler provide viewable compression metrics for devices. Figure 107 includes a view where the compression ratio of the device is shown. In this case the data is being compressed 5.0 to 1.
Figure 107. Data Reduction ratio for a PowerMax device
Using Solutions Enabler, it is also possible to obtain this information. Figure 108 shows how to list the data reduction metrics for an individual device that is part of a storage group with data reduction enabled.
Figure 108. Data Reduction detail in Solutions Enabler
Note: All supported data services, such as SnapVX, SRDF, vVols, and encryption (D@RE) are supported with data reduction.
Customer applications are one of the biggest factors that can impact data reduction ratios. For example, if the application has the ability to compress data before storing it, it is unlikely the PowerMax will compress it further. Oracle Compression is a good example of a feature of the Oracle Database that will benefit little from array data reduction. Similarly, software that encrypts data before storing it on the array cannot take advantage of the deduplication capability of array data reduction. Therefore, features such as these may guide placement of devices in storage groups without data reduction enabled.