Storage Modeller capacity calculation
Storage Modeller uses specific terms for different representations of the system capacity. A complete list of terms and their definition is available in
“General Glossary for IBM Storage Modeller”.
The following terms are specially relevant:
Raw capacity Reported capacity of the drives in the system before formatting or RAID is applied.
Usable capacity Capacity that is available for storing data on the system, pool, array, after formatting and RAID are applied.
Effective Capacity: Capacity that can be provisioned in a system or pool without running out of usable capacity, given the data reduction savings.
Storage Modeller calculates those capacities as follows:
▪Raw capacity is equal to the total of each drive or FCMs multiplied by their respective marketed gross capacity.
▪The usable capacity is the raw capacity minus the overhead for RAID formatting, metadata and capacity that could be reserved by the drive itself. The usable capacity is the amount of
non-reduced data that can be stored.
▪The effective capacity is calculated as the usable capacity multiplied by the data reduction ratio and represents the capacity effectively available to hosts.
Capacity calculation with IBM FlashCore Modules(FCM)
Storage Modeller precisely calculates raw and effective capacity. These calculations are aligned with the way that IBM accounts for FCM arrays, data reduction pools, garbage collection,
and other issues. You will sometimes observe differences with what would have been reported with former capacity configuration software.
System 9100 control enclosures can be equipped with IBM FlashCore Modules that provide automatic data compression.
These self-compressing drives present a logical capacity to the RAID controller. Storage Modeller will calculate an actual effective capacity that takes into account the pool settings,
array configuration (RAID type, Grouping and number of drives), the compression reduction and also what FCMs reserve internally (about 2 percent of the available capacity).
Example capacity calculations: Regular pool
versus Data Reduced pool
This section shows you how Storage Modeller calculates raw, usable, and effective capacity for two example systems. This section also shows the differences between a
Regular pool versus a Data Reduction pool (DRP), plus Deduplication.
The screen captures in
“Example capacity calculations as displayed in Storage Modeller”
show the completed Capacity calculations for two simple System 9500 devices. These devices share the following key characteristics:
▪One control enclosure
▪24 FlashCore Modules of 9.6TB
▪10+P+Q DRAID 6
The pool for our DRP example device differs in that we applied the following settings in the
Add Pool dialog box:
▪Pool Type:
Data Reduction Pool
(not
Regular pool)
▪Deduplication %: 30% (not 0%)
Table 30 Capacity calculations: Regular pool versus Data Reduced pool
|
Pool Type
|
Raw Capacity
(# of Drives X TB) |
Usable Capacity
(Raw Capacity) - (RAID)1
and also DRP-related deductions2
|
Effective
(Usable Capacity)
X (20% savings from compression3)
and also for DRP example
X (30% savings from deduplication4) |
|---|---|---|---|
|
Regular pool
|
24 x 9.6TB
= 230.39 TB = (209.54 TiB) |
230.39 TB (209.54TiB
- (5 drives dedicated to RAID)
= 182.35TB (165.85 TiB) |
182.35 TB (165.85 TiB) X 1.25 for compression
= 227.93 TB (207.31 TiB)
|
|
Data Reduction pool
(DRP)
|
230.39 TB (209.51 TiB)
- (RAID parity)
- (DRP-related deductions) = 161.10 TB (146.52 TiB) |
161.10 TB X 1.25 for compression = 201.38 TB X 1.43 for deduplication
= 287.68TB (261.65 TiB)
|
1 DRAID 6 and a 10+P+Q configuration affects a set of 24 drives as follows:
- 4 drives are used as parity drives (2 parity drives for each set of 10 drives).
- 1 drive is reserved as Spare capacity.
- Total number of drives dedicated to RAID6 is 5.
- For our examples, RAID reduces Raw Capacity by 38.4 TB (5 drives X 7.68 TB).
- 4 drives are used as parity drives (2 parity drives for each set of 10 drives).
- 1 drive is reserved as Spare capacity.
- Total number of drives dedicated to RAID6 is 5.
- For our examples, RAID reduces Raw Capacity by 38.4 TB (5 drives X 7.68 TB).
2 For DRP metadata and space needs, Storage Modeller reserves space (< 10%).
3 1.25 is the reciprocal of 80% (8/10). Multiplying by the reciprical is the same as dividing the Usable Capacity by 80%, which is the result of a 20% compression.
4 1.43 is the reciprocal of 70% (7/10). Multiplying by the reciprical is the same as dividing the Usable Capacity by 70%, which is the result of a 30% compression.
Example capacity calculations as displayed in Storage Modeller
This section shows how Storage Modeller displays the capacity calculations that are described in
Table 30:
▪Figure 351 on page 401
shows the calculation results for no DRP and DRP systems, side by side.
▪Figure 352 on page 401
shows the Capacity tab for the no-DRP example.
▪Figure 353 on page 402
shows the Capacity tab for the DRP example.
Figure 351 Capacity display for Regular pool (“noDRP”) versus Data Reduced pool (“DRP”)
Figure 352 Calculation example: pool capacity without DRP
Figure 353 Calculation example: pool capacity with DRP and deduplication