Planning Storage Pools

OceanStor 2810 V500R007 Basic Storage Service Configuration Guide for Block

This document is applicable to OceanStor 2810 V5. It describes the basic storage services and explains how to configure and manage them.

Rate and give feedback:

Huawei uses machine translation combined with human proofreading to translate this document to different languages in order to help you better understand the content of this document. Note: Even the most advanced machine translation cannot match the quality of professional translators. Huawei shall not bear any responsibility for translation accuracy and it is recommended that you refer to the English document (a link for which has been provided).

Planning Storage Pools

This section describes how to plan storage pools, including RAID levels.

Consider the following when selecting RAID levels:

Reliability
Read/Write performance
Disk utilization

Different RAID levels provide different reliability, read/write performance, and disk utilization, as described in Table 3-5 and Table 3-6.

Table 3-5 Reliability, read performance, and write performance of various RAID levels

RAID Level	Redundancy and Data Recovery Capability	Read Performance	Write Performance	Maximum Number of Allowed Faulty Disks
RAID 0	No data redundancy is provided and corrupted data cannot be recovered.	High	High	0
RAID 1	2D^a: relatively high 4D: high RAID 1 provides complete redundancy. If a CK fails, the mirror CK can be used to recover the data.	Relatively high	Relatively low	2D N (in a RAID 1 array consisting of 2N disks) with the best reliability performance 1 with the worst reliability performance 4D N+1 (in a RAID 1 array consisting of 2N disks) with the best reliability performance 3 with the worst reliability performance NOTE: With both data and mirroring chunks, the system delivers the best reliability performance, that is, can tolerate the concurrent failure of the most disks, in either of the following scenarios: All data chunks are faulty, but all of them have corresponding normal mirroring chunks. All mirroring chunks are faulty, but all of them have corresponding normal data chunks. With both data and mirroring chunks, the system delivers the worst reliability performance, that is, can tolerate the concurrent failure of the fewest disks when a data chunk and all its mirroring chunks are faulty.
RAID 3	Relatively high. Each CKG has one CK as the parity CK. Data on any data CK can be recovered using the parity CK. If two or more CKs fail, RAID 3 protection can no longer be provided.	High	Low	1
RAID 5	Relatively high. Parity data is distributed on different CKs. In each CKG, the parity data occupies space of a CK. RAID 5 allows the failure of only one CK. If two or more CKs fail, RAID 5 protection can no longer be provided.	Relatively high	Relatively high	1
RAID 6	High. Two groups of parity data are distributed on different CKs. In each CKG, the parity data occupies space of two CKs. RAID 6 allows two CKs to fail simultaneously. If three or more CKs fail, RAID 6 protection can no longer be provided.	Medium	Medium	2
RAID 10	Relatively high. RAID 10 allows multiple CKs to fail simultaneously. If a CK fails, the mirror CK can be used to recover the data. If a CK and its mirror CK fail simultaneously, RAID 10 protection can no longer be provided.	Relatively high	Relatively high	N (in a RAID 10 array consisting of 2N disks) with the best reliability performance 1 with the worst reliability performance NOTE: With both data and mirroring chunks, the system delivers the best reliability performance, that is, can tolerate the concurrent failure of the most disks, in either of the following scenarios: All data chunks are faulty, but all mirroring chunks are normal. All mirroring chunks are faulty, but all data chunks are normal. With both data and mirroring chunks, the system delivers the worst reliability performance, that is, can tolerate the concurrent failure of the fewest disks when a data chunk and its mirroring chunks are faulty.
RAID 50	Relatively high. Parity data is distributed on different CKs of each RAID 5 sub-group. In each RAID 5 sub-group, only one CK is allowed to fail. If two or more CKs of a RAID 5 sub-group fail simultaneously, RAID 50 protection can no longer be provided.	Relatively high	Relatively high	1
a: D indicates a data block.

Table 3-6 Typical disk utilization of various RAID levels

RAID Level	Typical Disk Utilization
RAID 0	100%
RAID 1	2D^a: about 50% 4D: about 25%
RAID 3	RAID 3 supports flexible configurations. Specifically, a RAID 3 policy allows data block and parity block policies ranging from 2D+1P to 13D+1P. The following examples show disk utilization of several configurations commonly used by RAID 3: 4D+1P^b: about 80% 2D+1P: about 66.67% 8D+1P: about 88.89%
RAID 5	RAID 5 supports flexible configurations. Specifically, a RAID 5 policy allows data block and parity block policies ranging from 2D+1P to 13D+1P. The following examples show disk utilization of several configurations commonly used by RAID 5: 2D+1P: about 66.67% 4D+1P: about 80% 8D+1P: about 88.89%
RAID 6	RAID 6 supports flexible configurations. Specifically, a RAID 6 policy allows data block and parity block policies ranging from 2D+2P to 26D+2P. The following examples show disk utilization of several configurations commonly used by RAID 6: 2D+2P: about 50% 4D+2P: about 66.67% 8D+2P: about 80% 16D+2P: about 88.89%
RAID 10	50%
RAID 50	(2D+1P) x 2: about 66.67% (4D+1P) x 2: about 80% (8D+1P) x 2: about 88.89%
a: D indicates a data block. b: P indicates a parity block. NOTE: For a flexibly configured RAID policy xD+yP, the disk utilization is [x/(x + y)] x 100%. If the RAID level of one storage tier is configured with flexible configuration first, this tier is the primary control tier that controls other tiers' RAID policies. When the RAID policy for the primary control tier is xD+yP, the RAID policy for other tiers must be: (xD+yP) multiplied by a factor of 1, 2, 4, or 8; or (xD+yP) divided by 1, 2, 4, or 8. For example, if the performance tier is the primary control tier and its RAID policy is 3D+1P, the RAID policy of other tiers must be 3D+1P, 6D+2P, or so on, and cannot be 4D+1P. If you want to change the current primary control tier, deselect this tier and select it again.

You are advised to use NL-SAS disks to create RAID 6 (26D+2P).

Table 3-7 describes the best practices for storage pool planning in video surveillance scenarios.

Table 3-7 Best practices for storage pool planning

Disk Quantity in a Disk Domain (M)	RAID Policy
M ≥ 30	RAID 6 (26D+2P)
28 ≤ M < 30	RAID 6 (24D+2P)
26 ≤ M < 28	RAID 6 (22D+2P)
18 ≤ M < 26	RAID 6 (14D+2P)
12 ≤ M < 18	RAID 6 (8D+2P)
8 ≤ M < 12	RAID 6 (4D+2P)
6 ≤ M < 8	RAID 6 (2D+2P)

The evaluation results of the local Huawei representative office or Huawei authorized distributor supersede the considerations in this section.

Translation

简体中文

Download

Updated: 2023-09-13

Document ID: EDOC1100138513

Downloads: 31

Average rating:

This Document Applies to these Products

Related Version

Digital Signature File

Digital Signature Authentication Mode

Enterprise

Huawei Cloud

Carrier

Consumer

Corporate

This document is applicable to OceanStor 2810 V5. It describes the basic storage services and explains how to configure and manage them.

Related Version

Related Documents

Digital Signature File