31 octobre 2025
admin-wp Huawei

IBM SAN Volume Controller(SVC) Gateway Array Heterogeneous Storage Takeover Testing with Huawei OceanStor Dorado All-Flash Storage

Logo

Axians Global

All Rights Reserved

Executive Summary

Axians Global (“Axians”) assessed the interoperability of IBM SAN Volume Controller(SVC) Gateway Array Heterogeneous Storage Takeover Huawei OceanStor Dorado All-Flash Storage .

In the assessment, Axians has determined that IBM SAN Volume Controller(SVC) will function with Huawei Storage for the following scenarios:

Test Scenario

Storage Involved

Protocol(s) Tested

Result

The Huawei storage is taken over by the IBM SAN Volume Controller (SVC) gateway Array. After the storage resources are pooled by the gateway, they are reallocated to hosts, simplifying management

Huawei OceanStor Dorado V700

FC

Passed

In this document, you will find details on the above test cases and the captured output. The procedures conducted in the tests are referenced from standard online documentation from IBM and Huawei.

1. Environment Configuration

1.1 Networking diagram

Figure 1.1 Networking Requirements for Two-Node IBM SVC Taking Over Four-Controller OceanStor Dorado 8000

1.2 IBM SVC Networking Requirements

According to the official documents and best practices of IBM SVC ,the storage network must meet the following requirements:

Each Huawei storage system can use up to 16 Fibre Channel ports to connect to IBM SVC. There are 4 Fibre Channel ports used to connect to IBM SVC in this Test.

Each port of the storage must be detected by all nodes of IBM SVC. Otherwise, the storage and MDisk are degraded, and an error is reported in the log indicating that repairing is required.

The storage’s Fibre Channel ports are dedicated to IBM SVC and cannot be shared with other hosts or used for other purposes. Port sharing is only functional during data migration and for a limited time.

When you configure zones for communication between IBM SVC nodes in the same system, up to 16 port logins are allowed from one node to any other node in the same fabric. Note that Fibre Channel ports connected to hosts or storage can also be used for communication between nodes

Note: Within each test configuration, a single zone comprises all initiator ports from an IBM SAN Volume Controller (SVC) node and one target port of an OceanStor Dorado V700 storage system, residing within the same Fibre Channel fabric

1.3 Hardware and Software Configuration

1.3.1 Storage Configuration

Table 1-1 Huawei storage configuration table

Name

Model

Version

Quantity

Gateway Array

IBM SAN Volume Controller(SVC)

7.8.1

1

Storage

OceanStor Dorado 8000

V700

1

1.3.2 Matching Hardware Configuration

Table 1-2 Hardware Configuration

Name

description

Usage

Quantity

RH_2288H_V3 Server

  • X86 architecture server

Installing Linux system for Host Server

1

SNS_3664 32G FC Switch

Network switch

Connecting two IBM SVC nodes as a cluster, networking between IBM SVC and Huawei storage, and the business network connecting IBM SVC to servers

2

1.3.3 Test Software and Tools

Table 1-3 Test Software and Tool List

Software Name

description

Version

Quantity

Operation System

Install the Linux Hosts

CentOS 8.0

1

Vdbench tool

Software for simulating business IO issuance

50407

1

2. Test Preparation

2.1 Scan initiators from IBM SVC on Huawei storage

2.2 OceanStor Dorado V700 Configuration Requirements

When creating a host on OceanStor Dorado V700, set the host OS to Linux and retain the default host access mode (load balancing).

OceanStor Dorado V700 uses the following method to create hosts.

Method : Use all initiators of IBM SVC to create a host.

Note:

OceanStor Dorado V700 storage systems use the active-active architecture.

SVC uses the WWNN of OceanStor Dorado V700 as the unique identifier of a controller. An OceanStor Dorado V700 storage system reports only one WWNN (not the WWPN of a Fibre Channel port), which is identified as a controller on IBM SVC.

2.3 Check the numbles and status of the paths

2.4 IBM SVC Comfiguration Requirements

When adding the MDisk to a pool on IBM SVC, you can select Flash, Enterprise, or NearLine for Tier based on service requirements. This parameter is for the Easy Tier function (similar to SmartTier of Huawei storage) of IBM SVC. When an OceanStor Dorado all flash storage system is used, select Flash.

When creating a volume on IBM SVC, you can select None, Thin-provisioned, or Compressed for Capacity saving, where Thin-provisioned is recommended.

If Thin-provisioned is selected, the volume is in the Online state after being created.

After volumes are created on IBM SVC, Cache mode is Enabled by default. OceanStor Dorado V700 is an all-flash storage system. According to the best practices of IBM SVC, you are advised to set Cache mode to Disabled

2.5 Create a host with the initiators from the test server on IBM SVC

3. Heterogeneous Storage Takeover Functional Testing

3.1 Huawei Storage Controller Identification

Prerequisites

  1. The gateway Array status is normal.
  2. The Huawei storage status is normal.
  3. The network has been set up according to the recommended networking diagram, and physical links are functioning properly.
  4. Switch configurations have been completed, and all initiators from Huawei storage to gateway array are online.

Test Procedure

  1. Create a 2TB Thin LUN on Huawei storage.
  2. Map the LUN from Huawei storage to the gateway storage.
  3. Scan for external storage on the gateway storage.
  4. Check the number of controllers on the Huawei storage device on the gateway storage.

Expected Result

  1. In step 4, the number of controllers stored on the gateway for Huawei storage devices is 1.

Test Result

  1. Step 1 Create a 2TB Thin LUN

2. Step 2 Map the LUN


3. Step 3 Scan for external storage on the gateway storage



4. Step 4 the number of controllers stored on the gateway for Huawei storage devices is 1


Test Conclusion

Passed

3.2 Huawei Storage WWN/SN Identification

Prerequisites

  1. The gateway Array status is normal.
  2. The Huawei storage status is normal.
  3. The network has been set up according to the recommended networking diagram, and physical links are functioning properly.
  4. Switch configurations have been completed, and all initiators from Huawei storage to gateway array are online.

Test Procedure

  1. Create a 2TB Thin LUN on Huawei storage.
  2. Map the LUN from Huawei storage to the gateway storage.
  3. Scan for external storage on the gateway storage.
  4. Check whether the gateway storage uses the WWN or SN of the Huawei storage device as the unique identifier, and compare it with the DeviceManager of Huawei storage.

Expected Result

  1. In step 4, the gateway stores whether the WWN or SN of the Huawei storage device is used as the unique identifier for recording, and it matches the information on Huawei Storage DeviceManager.

Test Result

  1. Step 1 Create a 2TB Thin LUN



2. Step 2 Map the LUN



3. Step 3 Scan for external storage on the gateway storage



4. Step 4 Check the gateway storage uses the WWN of the Huawei storage device as the unique identifier



compare it with the DeviceManager of Huawei storage, and it matches the information on Huawei Storage DeviceManager




Test Conclusion

Passed

3.3 PID Identification

Prerequisites

  1. The gateway Array status is normal.
  2. The Huawei storage status is normal.
  3. The network has been set up according to the recommended networking diagram, and physical links are functioning properly.
  4. Switch configurations have been completed, and all initiators from Huawei storage to gateway array are online.

Test Procedure

  1. Create a 2TB Thin LUN on Huawei storage.
  2. Map the LUN from Huawei storage to the gateway storage.
  3. Scan for disks on the gateway storage.
  4. Check the PID of the scanned Huawei storage device on the gateway storage.

Expected Result

  1. No errors are reported during the scan in Step 3, and the LUN mapped by Huawei storage is successfully detected on the gateway storage.
  2. In Step 4, the PID is XSG1.

Test Result

  1. Step 1 Create a 2TB Thin LUN

2. Step 2 Map the LUN



3. Step 3 Scan for external storage on the gateway storage



4. Step 4 ,the PID is XSG1


Test Conclusion

Passed

3.4 VID Identification

Prerequisites

  1. The gateway Array status is normal.
  2. The Huawei storage status is normal.
  3. The network has been set up according to the recommended networking diagram, and physical links are functioning properly.
  4. Switch configurations have been completed, and all initiators from Huawei storage to gateway array are online.

Test Procedure

  1. Create a 2TB Thin LUN on Huawei storage.
  2. Map the LUN from Huawei storage to the gateway storage.
  3. Scan for external storage on the gateway storage.
  4. Check the VID of the scanned Huawei storage device on the gateway storage.

Expected Result

  1. No errors are reported during the scan in Step 3, and the LUN mapped by Huawei storage is successfully detected on the gateway storage.
  2. In Step 4, the VID is HUAWEI.

Test Result

  1. Step 1 Create a 2TB Thin LUN



2. Step 2 Map the LUN



3. Step 3 Scan for external storage on the gateway storage



4. Step 4 ,the VID is HUAWEI


Test Conclusion

Passed

3.5 LUN WWN Identification

Prerequisites

  1. The gateway Array status is normal.
  2. The Huawei storage status is normal.
  3. The network has been set up according to the recommended networking diagram, and physical links are functioning properly.
  4. Switch configurations have been completed, and all initiators from Huawei storage to gateway array are online.

Test Procedure

  1. Create a 2TB Thin LUN on Huawei storage.
  2. Map the LUN from Huawei storage to the gateway storage.
  3. Scan for disks on the gateway storage.
  4. Check the WWN of the LUN on the gateway storage.

Expected Result

  1. In step 3, no errors are reported during scanning, and the LUNs mapped by Huawei storage are successfully detected on the gateway storage.
  2. In step 4, the gateway storage can recognize the WWN of the LUN, which is consistent with that on the Huawei storage.

Test Result

  1. Step 1 Create a 2TB Thin LUN



2. Step 2 Map the LUN,



3. Step 3 Scan for external storage on the gateway storage, no errors are reported during scanning, and the LUNs mapped by Huawei storage are successfully detected on the gateway storage



4. Step 4, Check the WWN of the LUN on the gateway storage



5. It is consistent with that on the Huawei storage


Test Conclusion

Passed

4. Heterogeneous LUN interoperability Testing

4.1 Minimum LUN Identification

Prerequisites

  1. The gateway Array status is normal.
  2. The Huawei storage status is normal.
  3. The network has been set up according to the recommended networking diagram, and physical links are functioning properly.
  4. Switch configurations have been completed, and all initiators from Huawei storage to gateway array are online.
  5. No LUNs have been mapped from Huawei storage to gateway storage.

Test Procedure

  1. Create a Thin LUN on Huawei storage with the minimum configurable capacity.
  2. Map the LUN from Huawei storage to the gateway storage.
  3. Scan for disks on the gateway storage.
  4. Check the capacity information of the scanned LUN on the gateway storage and compare it with the information on Huawei storage DeviceManager.

Expected Result

  1. In step 3, no errors are reported during scanning, and the LUNs mapped by Huawei storage are successfully detected on the gateway storage.
  2. In step 4, the capacity of the detected LUNs on the gateway storage is consistent with that on Huawei storage DeviceManager.

Test Result

  1. Step 1 the minimum configurable capacity is 512KB



2. Step 2,Map the LUN from Huawei storage to the gateway storage



3. In step 3, no errors are reported during scanning, and the LUNs mapped by Huawei storage are successfully detected on the gateway storage





4. In step 4, the capacity of the detected LUNs on the gateway storage is consistent with that on Huawei storage DeviceManager


Test Conclusion

Passed

4.2 Maximum LUN Identification

Prerequisites

  1. The gateway Array status is normal.
  2. The Huawei storage status is normal.
  3. The network has been set up according to the recommended networking diagram, and physical links are functioning properly.
  4. Switch configurations have been completed, and all initiators from Huawei storage to gateway array are online.
  5. No LUNs have been mapped from Huawei storage to gateway storage.

Test Procedure

  1. Create a Thin LUN on Huawei storage with the maximum configurable capacity.
  2. Map the LUN from Huawei storage to the gateway storage.
  3. Scan for disks on the gateway storage.
  4. Check the capacity information of the scanned LUN on the gateway storage and compare it with the information on Huawei storage DeviceManager.

Expected Result

  1. In step 3, no errors are reported during scanning, and the LUNs mapped by Huawei storage are successfully detected on the gateway storage.
  2. In step 4, the capacity of the detected LUNs on the gateway storage is consistent with that on Huawei storage DeviceManager.

Test Result

  1. Step 1 the Maximum configurable capacity is 256TB





2. Step 2,Map the LUN from Huawei storage to the gateway storage



3. In step 3, no errors are reported during scanning, and the LUNs mapped by Huawei storage are successfully detected on the gateway storage





4. In step 4, the capacity of the detected LUNs on the gateway storage is consistent with that on Huawei storage DeviceManager


Test Conclusion

Passed

4.3 Minimum and Maximum Host LUN ID Identification

Prerequisites

  1. The gateway Array status is normal.
  2. The Huawei storage status is normal.
  3. The network has been set up according to the recommended networking diagram, and physical links are functioning properly.
  4. Switch configurations have been completed, and all initiators from Huawei storage to gateway array are online.

Test Procedure

  1. Create two 2TB thin LUNs on Huawei storage and map them to the gateway storage. The host LUN IDs should be the minimum ID (0) and the maximum ID respectively.


The maximum host LUN ID varies depending on the model.

  1. Scan for disks on the gateway storage.
  2. Use the scanned LUNs to create pools/volumes on the gateway storage.

Expected Result

  1. No errors are reported during the scan in Step 2, and the LUNs mapped by Huawei storage are successfully scanned on the gateway storage.
  2. volumes are successfully created in Step 3.

Test Result

  1. Step1, The host LUN IDs should be the minimum ID (0) and the maximum ID(4095)



2. Step2, No errors are reported during the scan, and the LUNs mapped by Huawei storage are successfully scanned on the gateway storage







3. Step3,volumes are successfully created








Test Conclusion

Passed

4.4 Map the taken-over LUN to the host for read/write testing

Prerequisites

  1. The gateway Array status is normal.
  2. The Huawei storage status is normal.
  3. The network has been set up according to the recommended networking diagram, and physical links are functioning properly.
  4. Switch configurations have been completed, and all initiators from Huawei storage to gateway array are online.

Test Procedure

  1. Create two LUNs on Huawei storage and map them to the gateway storage.
  2. Scan for disks on the gateway storage, create pools/volumes using the detected LUNs, and map them to the host.
  3. On the host, create a file system on one volume, write files to it, and use a tool to continuously read and write data on the other volume until the test case ends. The I/O model is as follows: 8KB I/O, read/write ratio of 7:3, fully random, 8 concurrent operations, iorate=1000. Record the bandwidth, IOPS, and latency of the tool.
  4. observe the operation status of the tool.

Expected Result

  1. In Step 5, files in the file system on the host can be read and written normally, the tool runs without interruption, and there is no significant change in performance.

Test Result

  1. Step1,Create two
LUNs on Huawei storage and map them to the gateway storage.

2. Step 2,Scan for disks on the gateway storage, create pools/volumes using the detected LUNs, and map them to the host






3. On the host, create a file system on one volume, write files to it, and use a tool to continuously read and write data on the other volume until the test case ends








4. Step5, files in the file system on the host can be read and written normally, the tool runs without interruption, and there is no significant change in performance.


Test Conclusion

Passed

5. Reliability Testing

5.1 Switch Networking – Long-Term Disconnection and Recovery of a Single Link on Huawei Storage Side

Prerequisites

  1. The gateway array status is normal.
  2. The Huawei storage status is normal.
  3. The gateway storage has taken over the Huawei storage and created a volume, which has been mapped to the host.

Test Procedure

  1. On the host side, use Vdbench to read and write the volumes mapped by the gateway storage until the test case ends. The I/O model is: 8KB I/O, read/write ratio of 7:3, fully random, 8 concurrent operations, iorate=1000. Record the bandwidth, IOPS, and latency of the tool.
  2. Check the traffic on the remote link between the gateway storage and the switch.
  3. Disconnect one remote link between the Huawei storage and the switch that has I/O traffic, and observe the Vdbench operation status and the traffic on the remote link.
  4. After one minute, restore the remote link and observe the Vdbench operation status and the traffic on the remote link.
  5. Repeat steps 3-4 for a total of three tests.

Expected Result

  1. In step 3, Vdbench runs without interruption. The traffic on the disconnected remote link is switched to other links. The switchover time has been recorded, and the duration of Vdbench IO fluctuation/return to zero has also been recorded.
  2. In step 4, the duration of Vdbench IO fluctuation/return to zero has also been recorded.

Test Result

  1. Step1, On the host side, use Vdbench to read and write the volumes mapped by the gateway storage until the test case ends







2. Step2, Check the traffic on the remote link between the gateway storage and the switch









3. Step3,Vdbench runs without interruption. The traffic on the disconnected remote link is switched to other links. The switchover time has been recorded, and the duration of Vdbench IO fluctuation/return to zero has also been recorded(10 second)










4. Step4, the duration of Vdbench IO fluctuation/return to zero has also been recorded







5. Repeat steps 3-4 for a total of three tests

Disconnect one remote link between the Huawei storage and the switch that has I/O traffic, and observe the Vdbench operation status and the traffic on the remote link











After one minute, restore the remote link and observe the Vdbench operation status and the traffic on the remote link









Disconnect one remote link between the Huawei storage and the switch that has I/O traffic, and observe the Vdbench operation status and the traffic on the remote link









After one minute, restore the remote link and observe the Vdbench operation status and the traffic on the remote link












Test Conclusion

Passed

5.2 Switch Networking – Gateway Array Side Single Link Disconnection and Recovery for an Extended Period of Time

Prerequisites

  1. The gateway array status is normal.
  2. The Huawei storage status is normal.
  3. The gateway storage has taken over the Huawei storage and created a volume, which has been mapped to the host.

Test Procedure

  1. On the host side, use Vdbench to read and write the volumes mapped by the gateway storage until the test case ends. The I/O model is: 8KB I/O, read/write ratio of 7:3, fully random, 8 concurrent operations, iorate=1000. Record the bandwidth, IOPS, and latency of the tool.
  2. Check the traffic on the remote link between the gateway storage and the switch.
  3. Disconnect one remote link between the gateway storage and the switch that has I/O traffic, and observe the Vdbench operation status and the traffic on the remote link.
  4. After one minute, restore the remote link and observe the Vdbench operation status and the traffic on the remote link.
  5. Repeat steps 3-4 for a total of three times.

Expected Result

  1. In step 3, Vdbench runs without interruption, and the I/O fluctuation/return-to-zero duration of Vdbench and the remote link has been recorded.

Test Result

  1. Step3,Vdbench runs without interruption,and the I/O fluctuation/return-to-zero duration of Vdbench and the remote link has been recorded


First Test:












After one minute, restore the remote link and observe the Vdbench operation status and the traffic on the remote link






Second Test:










After one minute, restore the remote link and observe the Vdbench operation status and the traffic on the remote link






Thind Test










After one minute, restore the remote link and observe the Vdbench operation status and the traffic on the remote link





Test Conclusion

Passed

5.3 A controller was abnormally removed and then restored.

Prerequisites

  1. The gateway array status is normal.
  2. The Huawei storage status is normal.
  3. The gateway storage has taken over the Huawei storage and created a volume, which has been mapped to the host.

Test Procedure

  1. Use the Vdbench test tool on the host to perform continuous read and write operations on the volume until the test case ends. The I/O model is: 8KB I/O, read/write ratio of 7:3, fully random, 8 concurrent operations, iorate=1000. Record the bandwidth, IOPS, and latency of the tool.
  2. Remove one controller from the Huawei storage system and observe the operation status of the Vdbench tool.
  3. Reinsert the controller and wait for the environment to return to normal.
  4. Remove the other controller from the gateway storage system and observe the operation status of the Vdbench tool.
  5. Reinsert the controller and wait for the environment to return to normal.

Expected Result

  1. In Step 2, the Vdbench IO fluctuation/reset status has been recorded.
  2. In Step 3, the environment can return to normal; if manual intervention is required, please record it.
  3. In Step 4, the Vdbench IO fluctuation/reset status has been recorded.
  4. In Step 5, the environment can return to normal; if manual intervention is required, please record it.
  5. Throughout the entire test process, Vdbench runs without interruption.

Test Result

  1. Step1, Use the Vdbench test tool on the host to perform continuous read and write operations on the volume until the test case ends





2. Step2, the Vdbench IO fluctuation/reset status has been recorded





3. Step3, the environment can return to normal













4. Step 4, the Vdbench IO fluctuation/reset status has been recorded





Step 5, the environment can return to normal













Throughout the entire test process, Vdbench runs without interruption

Test Conclusion

Passed

5.4 Abnormal Removal and Recovery of Interface Card

Prerequisites

  1. The gateway array status is normal.
  2. The Huawei storage status is normal.
  3. The gateway storage has taken over the Huawei storage and created a volume, which has been mapped to the host.

Test Procedure

  1. Use the Vdbench test tool on the host to perform continuous read and write operations on the volume until the test case ends. The I/O model is: 8KB I/O, read/write ratio of 7:3, fully random, 8 concurrent operations, iorate=1000. Record the bandwidth, IOPS, and latency of the tool.
  2. Remove one interface card with I/O traffic from the Huawei storage system and observe the operation status of the Vdbench tool.
  3. Reinsert the interface card and wait for the environment to return to normal.

Expected Result

  1. In Step 2, the Vdbench IO fluctuation/reset status has been recorded.
  2. In Step 3, the environment can return to normal. If manual intervention is required, please record it.
  3. Throughout the entire test process, Vdbench runs without interruption.

Test Result

  1. Step1, Use the Vdbench test tool on the host to perform continuous read and write operations on the volume until the test case ends





2. Step2, the Vdbench IO fluctuation/reset status has been recorded












3. Step3, the environment can return to normal













Throughout the entire test process, Vdbench runs without interruption

Test Conclusion

Passed

5.5 Huawei Storage System Abnormal Power Failure and Recovery

Prerequisites

  1. The gateway array status is normal.
  2. The Huawei storage status is normal.
  3. The gateway storage has taken over the Huawei storage and created a volume, which has been mapped to the host.

Test Procedure

  1. Create a file system on the volume on the host and write file data.
  2. Disconnect all power supplies of the Huawei storage system.
  3. Check the status of the Huawei storage device on the gateway storage.
  4. Restore power and wait for the device to return to normal operation.
  5. Check the status of the Huawei storage device on the gateway storage.
  6. Verify the data files on the host.

Expected Result

  1. In Step 3, the Huawei storage device on the gateway storage is in a faulty state.
  2. In Step 5, the status of the Huawei storage device on the gateway storage returns to normal.
  3. In Step 6, there is no data loss on the host.

Test Result

  1. Step1, Use the Vdbench test tool on the host to perform continuous read and write operations on the volume until the test case ends













    2. Check the md5 value of the file data which was write into the volume



    [root@localhost vdb.1_1.dir]# md5sum /mnt/filetest2/vdb.1_1.dir/vdb_f0000.file

    da1f618fe06f5db2f89710025af59b79 /mnt/filetest2/vdb.1_1.dir/vdb_f0000.file

    [root@localhost vdb.1_1.dir]# md5sum /mnt/filetest2/vdb.1_1.dir/vdb_f0001.file

    5c7da67f75bfb6d301c617031a730490 /mnt/filetest2/vdb.1_1.dir/vdb_f0001.file

    [root@localhost vdb.1_1.dir]# md5sum /mnt/filetest2/vdb.1_1.dir/vdb_f0002.file

    9a4ca3521110a94b6cd71e0db9554223 /mnt/filetest2/vdb.1_1.dir/vdb_f0002.file

2. Step3,the Huawei storage device on the gateway storage is in a Degraded state



3. Step5, the status of the Huawei storage device on the gateway storage returns to normal







4. Step6，No data loss in the filesystem



All md5 value of the files are same



vdb_f0000.file vdb_f0001.file vdb_f0002.file vdb_f0003.file vdb_f0004.file vdb_f0005.file vdb_f0006.file vdb_f0007.file vdb_f0008.file vdb_f0009.file

[root@localhost vdb.1_1.dir]# md5sum /mnt/filetest2/vdb.1_1.dir/vdb_f0000.file

da1f618fe06f5db2f89710025af59b79 /mnt/filetest2/vdb.1_1.dir/vdb_f0000.file

[root@localhost vdb.1_1.dir]# md5sum /mnt/filetest2/vdb.1_1.dir/vdb_f0001.file

5c7da67f75bfb6d301c617031a730490 /mnt/filetest2/vdb.1_1.dir/vdb_f0001.file

[root@localhost vdb.1_1.dir]# md5sum /mnt/filetest2/vdb.1_1.dir/vdb_f0002.file

9a4ca3521110a94b6cd71e0db9554223 /mnt/filetest2/vdb.1_1.dir/vdb_f0002.file

Test Conclusion

Passed

6. Reference

6.1 IBM SAN Volume Controller Document Resources

https://www.ibm.com/products/san-volume-controller