Administrator Guide

SAS, for example), and in the same tier, but can have different capacities. ADAPT is shown as a RAID level in the management
interfaces.
ADAPT disk groups use all available space to maintain fault tolerance, and data is spread evenly across all of the disks. When
new data is added, new disks are added, or the system recognizes that data is not distributed across disks in a balanced way, it
moves the data to maintain balance across the disk group.
Reserving spare capacity for ADAPT disk groups is automatic since disk space dedicated to sparing is spread across all disks
in the system. In the case of a disk failure, data will be moved to many disks in the disk group, allowing for quick rebuilds and
minimal disruption to I/O.
The system will automatically default to a target spare capacity that is the sum of the largest two disks in the disk group, which
is large enough to fully recover fault tolerance after loss of any two disks in the disk group. The actual spare capacity value can
change depending on the current available spare capacity in the disk group. Spare capacity is determined by the system as disks
are added to a disk group, or when disk groups are created, expanded, or rebalanced. For more information, see the topic about
the add disk-group command in the Dell EMC PowerVault ME4 Series Storage System CLI Guide.
ADAPT disk groups can be expanded to either replenish current target spare capacity or to increase usable capacity. For more
information, see Expanding a disk group on page 88.
A system using ADAPT disk groups cannot be downgraded to a system that does not support ADAPT.
About SSDs
The use of SSDs (solid-state drives) can greatly enhance the performance of a system. Since the SSDs do not have moving
parts, data that is random in nature can be accessed much faster. You can use SSDs for virtual disk groups. When combined
with virtual disk groups that consist of other classes of disks, improved read and write performance is possible through
automated tiered storage. Alternatively, you can use one or two SSDs in read-cache disk groups to increase read performance
for pools without a Performance tier. The application workload of a system determines the percentage of SSDs of the total disk
capacity that is needed for best performance.
For more information about automated tiered storage, see About automated tiered storage on page 24. For more information
on read-cache disk groups, see Read-cache disk groups on page 16. For information about using SSDs in all disk groups, see
All-flash array on page 19.
Gauging the percentage of life remaining for SSDs
An SSD can be written and erased a limited number of times. Through the SSD Life Left disk property, you can gauge the
percentage of disk life remaining. This value is polled every 5 minutes. When the value decreases to 20%, an event is logged
with Informational severity. This event is logged again with Warning severity when the value decreases to 5%, 2% or 1%, and
0%. If a disk crosses more than one percentage threshold during a polling period, only the lowest percentage will be reported.
When the value decreases to 0%, the integrity of the data is not guaranteed. To prevent data integrity issues, replace the SSD
when the value decreases to 5% of life remaining.
You can view the value of the SSD Life Left property through the Disk Information panel. In the front view of the enclosure in
the System topic, hover the cursor over any disk to view its properties. You can also view the Disk Information panel through
the Pools topic. Select the pool for the disk group in the pools table, select the disk group in the Related Disk Groups table, and
then hover the cursor over the disk in the Related Disks table.
All-flash array
The all-flash array feature, enabled by default, allows systems to run exclusively with disk groups that consist of SSDs, providing
the ability to have a homogeneous SSD-only configuration. Systems using an all-flash array have one tier that consists solely of
SSDs. If a system includes disk groups with spinning disks, the disk groups must be removed before the all-flash array feature
can be used.
If you are using SSDs and spinning disks and the first disk group is provisioned with spinning disks, then the system can be
provisioned to use spinning disks in virtual disk groups and use SSDs either in virtual disk groups or as read cache.
Getting started
19