No relevant resource is found in the selected language.

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>Search

Reminder

To have a better experience, please upgrade your IE browser.

upgrade

HyperClone Feature Guide for Block

OceanStor V5 Series V500R007

This document is applicable to OceanStor 5110 V5, 5110F V5, 5300 V5, 5300F V5, 5500 V5, 5500F V5, 5600 V5, 5600F V5, 5800 V5, 5800F V5, 6800 V5, 6800F V5, 18500 V5, 18500F V5, 18800 V5, and 18800F V5. This document describes the implementation principles and application scenarios of the clone feature. Also, it explains how to configure and manage clones.
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).
Configuration Example

Configuration Example

This chapter provides an example to explain how to configure the LUN clone feature in a typical application scenario. The configuration example includes requirement analysis, configuration planning, configuration operations, and configuration verification.

Performing Data Analysis Based on LUN Clone

This configuration example explains how to perform data analysis based on LUN clone without affecting the production service performance.

Requirement Analysis

This section analyzes the requirements of an enterprise that performs data analysis using the LUN clone feature.

Requirement Source

The enterprise has built an information system to store transaction data in real time and analyze the transaction data. The information system consists of a production cluster, a background analysis cluster, a storage system, and a maintenance terminal. The production cluster is composed of three application servers. The background analysis cluster is composed of two application servers. The storage system provides space to store service data. The maintenance terminal manages the storage system. Application servers are connected to the storage system through two Fibre Channel switches. Figure 2-2 shows the network diagram.

Figure 2-2 Network diagram

Table 2-4 describes the functions and characteristics of each cluster.

Table 2-4 Functions and characteristics of each cluster

Cluster

Function

Operating Time

I/O Characteristic and Requirement

Production cluster

This cluster carries the production service, stores transaction data, and provides query functions.

24/7 hours

The I/O traffic is heavy, requiring low latency and high performance.

Background analysis cluster

This cluster carries the data analysis service, such as analyzing transaction data in the background and outputting reports rapidly.

One to three times per month, irregularly triggered as required

The I/O traffic is heavy during data analysis and testing, requiring low latency and high performance.

Multiple LUNs are created in the storage system. Each LUN stores one month of the production cluster's service data. Table 2-5 describes the configuration of the storage system.

Table 2-5 Configuration of the storage system

Disk Configuration

LUN

Capacity

Mapped Host

48 x 600 GB SAS disks

  • ...
  • LUN_201304
  • LUN_201305
  • LUN_201306
  • ...

1 TB per LUN

  • Production cluster
  • Background analysis cluster

In the initial phase, the amount of transaction data is relatively small. Therefore, concurrent running of the production service and the data analysis service does not affect each other's performance greatly.

As the business keeps growing, transaction data access becomes increasingly intensive. Each time the data analysis service is running, the transaction data access latency increases greatly because the data analysis service contends with the production service for data from the same data source (the same LUN in the storage system). As a result, when customers query transaction information from the production cluster, the system response is significantly slow, adversely affecting the enterprise's business.

In addition, the production service consumes a lot of storage resources, making the storage system a bottleneck to the performance of the data analysis service. As a result, data analysis reports cannot be generated on time.

Facing the preceding problems, the enterprise needs a solution that ensures the performance of the production service and the data analysis service while minimizing costs. The specific requirements are as follows:

  • Requirement 1: The production service and the data analysis service do not have a noticeable impact on each other's performance.
  • Requirement 2: Data provided for the data analysis service must be synchronized in real time with the data generated by the production service to ensure that the latest data is accessed by the data analysis service.
  • Requirement 3: Data analysis is performed three times at most per month.
  • Requirement 4: New devices are not purchased in order to reduce costs.
Requirement Analysis and Conclusion
  • Feature selection analysis

    To prevent the production service and the data analysis service from affecting each other's performance, the enterprise must enable their data to be independent. To meet this need, the enterprise must create a copy of production data in the storage system. The storage system provides the following features related to data copies: snapshot, LUN copy, remote replication, and clone. Table 2-6 analyzes the extent to which each feature meets the enterprise's requirements.

    Table 2-6 Analysis of the extent to which each feature meets the enterprise's requirements

    Requirement

    Snapshot

    LUN Copy

    Remote Replication

    Clone

    Services do not have a noticeable impact on each other's performance.a

    Not supported (The pointer mechanism of the snapshot feature affects the performance of source LUNs.)

    Supported

    Supported

    Supported (After a clone is split, the primary LUN's performance is not affected.)

    Data is synchronized in real time.

    Supported

    Not supported (Copying data must be initiated manually.)

    Supported

    Supported (Real-time data synchronization is implemented before a clone is split.)

    Data analysis is performed three times per month.

    Supported

    Supported

    Supported

    Supported (as long as enough secondary LUNs are configured)

    The expenditure is low. No extra storage system needs to be purchased.

    Supported

    Supported

    No (At least one extra storage system must be purchased.)

    Supported (No extra storage system needs to be purchased.)

    a: The impact refers to that caused by the running of the data analysis service.

    The conclusion drawn from Table 2-6 is that only clone fully meets all the requirements.

  • Configuration policy analysis
    • Clone can be configured in the existing storage system without adding hardware or changing the live network.
    • The enterprise requires that the production service and the data analysis service do not have a noticeable impact on each other's performance. Therefore, the primary LUN (where production data resides) and secondary LUN (which is provided for the data analysis service) of a clone must be created in different disk domains. If the primary LUN and the secondary LUN are created in the same disk domain, the performance levels of the two LUNs are affected by the overall performance bottleneck of the disk domain, preventing clone advantages from being brought into full play.
    • The enterprise requires that real-time data synchronization be implemented between the production service and the data analysis service. Therefore, the clone must stay in the synchronized state and cannot be split until the data analysis service starts.
    • The enterprise performs data analysis three times at most per month. Therefore, three data copies must be prepared each month. The following schemes are available:
      • Create one secondary LUN and use it in each data analysis. After the LUN is used each time, synchronize it with the primary LUN again for the next data analysis.
      • Create three secondary LUNs and use one of them in each data analysis. After a secondary LUN is used, delete it.

        Data analysis is triggered irregularly. Therefore, the secondary LUN provided for data analysis must be available at any time. In scheme 1, after the secondary LUN is used in a data analysis, it must be synchronized with the primary LUN again for data consistency. The synchronization process takes some time, and the secondary LUN is unavailable during this process. Therefore, only scheme 2 is suitable.

    • The enterprise wants to reduce costs. Therefore, disk space consumption should be minimized. To meet this requirement, you are advised to delete all the secondary LUNs created in the previous month at the beginning of the following month regardless of whether they have been used before.
Configuration Planning

This section describes how to plan LUN clone configuration for data analysis.

It is inferred from the analysis conclusion that the clone feature enables the enterprise to perform data analysis without compromising the production service performance. Figure 2-3 shows the configuration roadmap designed to meet the enterprise's requirements.

Figure 2-3 Configuration roadmap

Table 2-7 describes the configuration operations and planning items involved in the configuration example.

NOTE:

Retain the default values of the parameters that are not mentioned in Table 2-7.

Table 2-7 Configuration operations and planning items

Configuration Roadmap

Major Operation and Purpose

Planning Item

Parameter

Planned Value

Preparing for the use of clone

  • Check the license file to ensure that it grants the permission to use the clone feature.
  • Create a storage pool and LUNs to store data involved in clone.

Parameters about the disk domain where the secondary LUNs residea

Name

Domain_Analysis

Disk Type

Performance Tier (SAS): 9b

Hot Spare Policy

High

Parameters about the storage pool where the secondary LUNs reside

Name

Pool_Analysis

Disk Domain

Domain_Analysis

Storage Medium

  • Performance tier (SAS)
  • Capacity: 3 TB

Parameters about the secondary LUNs

Name

  • LUN_201304_ana_01
  • LUN_201304_ana_02
  • LUN_201304_ana_03

Capacity

500 GB per LUN

Owning Controller

The same as that of the primary LUN

Configuring and managing a clone

  • Create a clone and set up a pair relationship between the LUN used by the production service (primary LUN) and each of the LUNs used by data analysis service (secondary LUNs).
  • Before starting the data analysis service, split a pair to make a secondary LUN available.

Parameters about the clone

Name

clone_for_analysis

Primary LUN Name

LUN_201304c

Secondary LUN Name

  • LUN_201304_ana_01
  • LUN_201304_ana_02
  • LUN_201304_ana_03

Copy Speed

Mediumd

Enable initial synchronization

Yes

Using secondary LUNs

Create a mapping view and map the secondary LUNs that are split from pairs to the data analysis cluster.

Parameters about the mapping view

Name

map_for_analysis

Mapped host

The hosts in the data analysis cluster are as follows:

  • ana_01
  • ana_02

Mapped LUN

Each time you need to perform data analysis, split a pair and add the secondary LUN to the mapping view. After performing data analysis, delete the secondary LUN. Available secondary LUNs are as follows:

  • LUN_201304_ana_01
  • LUN_201304_ana_02
  • LUN_201304_ana_03

LUN group that the secondary LUNs belong to

Name

LG_analysis

Host group that the data analysis cluster belongs to

Name

HG_analysis

a: To prevent performance from being affected, you are advised to create LUNs to be used by the data analysis service in a disk domain different from that where the LUNs used by the production service reside.

b: Three 1 TB LUNs need to be created in this storage pool. Therefore, the storage pool is required to provide at least 3 TB free space. RAID parity data, hot spare space, and storage pool metadata also need to be considered. Therefore, the storage pool must have at least 5 TB free space. Nine 600 GB SAS disks can be selected to meet the requirement.

c: In this example, the data amount generated in April 2013 is used as a reference. LUNs for the following months can be configured by referring to this planning example.

d: During the initial synchronization, a higher synchronization speed leads to a greater impact on the performance of the primary LUN. To strike a balance between the synchronization speed and the performance impact, you are advised to set Copy Speed to Medium.

Configuration Operations

You can configure the LUN clone feature on DeviceManager to provide data sources for data analysis.

Prerequisites

All configuration items have been planned.

Procedure
  1. Check the license file.

    A valid clone license must exist in the storage system.

    Operation path: Setting > License Management

  2. Create storage resources.

    1. Create a disk domain for the clone.

      Operation path: Provisioning > Disk Domain > Create

    2. Create a storage pool for the clone.

      Operation path: Provisioning > Storage Pool > Create

    3. View the owning controller of the primary LUN.

      Operation path: Choose Provisioning > LUN, select the primary LUN LUN_201304, and choose Properties > Advanced.

    4. In the storage pool created in 2.b, create three secondary LUNs for the clone.
      NOTE:

      The secondary LUNs must belong to the same owning controller as that of the primary LUN.

      Operation path: Provisioning > LUN > Create

  3. Create a clone and add the three secondary LUNs.

    Operation path: Data Protection > Clone > Create

    After a clone is created, an initial synchronization is implemented to keep data on the secondary LUNs the same as that on the primary LUN. Then, data is synchronized from the primary LUN to the secondary LUNs in real time to ensure that the secondary LUNs always store the latest data.

  4. Split a pair.

    When you need to start the data analysis service, split a pair.

    NOTE:

    Perform the split after the pair synchronization completes (when the Data Status is Consistent). A split performed when the Data Status is Synchronizing will lead to a data inconsistency between the primary LUN and the secondary LUN.

    On the Clone page, right-click the clone named clone_for_analysis and choose Split from the shortcut menu. Then, split the pair to which secondary LUN LUN_201304_ana_01 belongs.

    After the pair is split, LUN_201304_ana_01 serves as a data source for data analysis.

  5. Use the secondary LUN split from the pair.

    Map the secondary LUN to the data analysis cluster.

    1. Create a LUN group and add secondary LUN LUN_201304_ana_01 to it.

      Operation path: Provisioning > LUN > LUN Group > Create.

    2. Create a host group and add hosts ana_01 and ana_02 of the data analysis cluster to it.

      Operation path: Provisioning > Host > Host Group > Create.

    3. Create a mapping view and map LUN_201304_ana_01 to all hosts in the data analysis cluster.

      Operation path: Provisioning > Mapping View > Create.

    4. On the hosts that belong to the data analysis cluster, discover the mapped secondary LUN and start the data analysis service with the LUN as the data source.

Follow-up Procedure
  • After the data analysis service is complete, you are advised to delete the secondary LUN used by the service so that storage space is released for creation of new secondary LUNs.
  • Three secondary LUNs have been created for the three possible data analysis activities in the current month. When you need to start the data analysis service again, split another clone pair by repeating 4 and 5 to make another secondary LUN available for data analysis.
  • From the next month on, perform 2.d to 5 at the beginning of each month to create secondary LUNs for the data analysis service.
Configuration Verification

This section describes how to verify whether the enterprise's problems are resolved from three aspects: performance, real-time data synchronization, and the number of available data sources.

Procedure
  1. Check whether the mutual performance impact between the production service and the data analysis service is eliminated.

    Monitor the IOPS and latency of the production cluster and the data analysis cluster. Then, compare the monitoring results with those obtained before the use of clone to determine whether the IOPS and latency are improved.

    • If yes, clone has resolved the performance impact problem.
    • If no, clone has not resolved the performance impact problem. You are advised to check whether the configuration steps are performed correctly and whether the entire system has any other bottlenecks.

      Do not map the primary and secondary LUNs that are split from a pair to the same Windows-based host for data consistency verification. This practice may cause the operating system of the host to break down.

  2. Check whether the LUN used by the data analysis service is synchronized in real time with the LUN used by the production service.

    1. Split a clone pair and record the split time.
    2. Map the secondary LUN to the data analysis cluster.
    3. On a host that belongs to the data analysis cluster, check whether the data on the secondary LUN is the same as that on the primary LUN at the time when the pair was split. For example, you can check the time stamp.
      • If yes, clone meets the real-time data synchronization requirement.
      • If no, clone fails to meet the real-time data synchronization requirement. You are advised to check whether the configuration steps are performed correctly.

  3. Check whether there are three available data sources.

    In the preceding configuration process, one secondary LUN proves to be available. In this step, check whether the remaining two secondary LUNs are available.

    1. Repeat 4 and 5 described in Configuration Operations to check whether the remaining secondary LUNs are available.
    2. Repeat Configuration Verification to check whether the remaining secondary LUNs are available.
    3. If both secondary LUNs are available, the number of available data sources meets the requirement.

Translation
Download
Updated: 2019-07-11

Document ID: EDOC1000181501

Views: 22292

Downloads: 222

Average rating:
This Document Applies to these Products
Related Version
Related Documents
Share
Previous Next