Guide to Checking for Risks in HyperMetro Configuration
A total of 15 risks exist in HyperMetro configuration. Among which, 14 risks can be identified using SmartKit, while the rest one needs to be identified manually.
1 SmartKit Inspection
1. Download SmartKit of the latest version (SmartKit V2R5C00RC3 or later) from the support website.
Link: http://support.huawei.com/enterprise/en/cloud-storage/oceanstor-toolkit-pid-8576706/software
For OceanStor 18000 V3, download SmartKit of the latest version and install it in the SVP.
2. Start SmartKit and activate it as a Huawei or channel engineer. On the homepage, choose MyTools > Routine Maintenance > Inspection.

3. On the Welcome page, select Routine inspection and click Next.

4. On the Select Devices page, click Add Device. In the displayed dialog box, add the following devices:
l Local storage array with HyperMetro
l Remote storage array with HyperMetro
l Hosts that use HyperMetro LUNs

For V2R5C00RC1, hosts are covered under inspection in HyperMetro scenarios. Configuration on hosts to which the storage arrays in a HyperMetro pair are mapped can be checked (such as PDL configuration and ALUA configuration about using native multipathing software on ESXi hosts).
If the hosts are not added, their HyperMetro configuration cannot be checked.

l No debugging password is required for V3 storage arrays.
l When adding a FusionSphere host, select Linux for the host, set the IP address to the IP address of the CNA node instead of that of the BMC network port, use the gandalf account, and select Switch to user root.
If an added storage array is in a HyperMetro relationship, the
icon is displayed next to the storage array model. When you move your cursor on the icon, a message pops up showing the remote storage array. If one of the storage arrays in a HyperMetro pair is selected, the other one is selected automatically.
Hosts are covered under inspection, but the inspection cannot be independently performed for hosts.
If the local storage array in a HyperMetro pair is added but the remote one is not added, a message is displayed prompting you to add the remote storage array and the related host.
For the check items, a new parent node (HyperMetro information consistency) is added. Under this parent node, the configuration of both storage arrays in a HyperMetro pair is checked.
After the hosts are added, they are displayed in Inspection devices.
5. Specify the path for saving the inspection report and click Next.
6. On the Start Inspection page, click the device to view details about its inspection progress. After the inspection is complete, refer to the rectification suggestions for failed check items. For any trouble, contact R&D support engineers or TAC/GTAC for help.
7. Click Open the res… View the rep.. to view the inspection report.
2 Manual Confirmation
The disk used by a quorum server cannot be a disk mapped from the storage array in a HyperMetro pair. Otherwise, the quorum server cannot work if the storage array becomes faulty.
VMware and FusionSphere are used as examples.
VMware ESX
1. Log in to the VMware vSphere client, find the VM running the quorum server and confirm the datastore used by the VM.
2. Choose Configuration > Storage, and then check the device information about the datastore.
If the device information contains the keyword "Huawei" and the WWN belongs to a storage array in a HyperMetro pair, the disk used by the quorum server is mapped from the storage array in a HyperMetro pair. The risk can be confirmed and the disk needs to be migrated to a local one.
FusionCompute
1. Log in to the FusionCompute home page, find the VM running the quorum server on the VM and Template page.
2. Click the VM name and check the datastore to which the server belongs.
3. Choose Storage > Data Store and check the type of the datastore.
If Type is Local, there is no risk. Otherwise, migrate the disk to a local one.
3 Appendix: FAQs
3.1 What Should I Do If the Management Port of the Storage Array Is Not Connected to the Network?
Advise the customer to connect the management port of the storage array to the network for routine maintenance. Alternatively, use a small switch to connect the two storage arrays, and then connect your laptop to the switch for inspection.
3.2 What Should I Do If a Host Fails to Be Added for Inspection?
The management network of a storage array is isolated from the service network of the related host and host services cannot be accessed due to confidentiality, so the host cannot be added during the inspection of the storage array. You can resolve this problem using the following methods:
Method 1: Perform inspection for the storage array and hosts separately. Specifically, use the inspection tool to inspect the storage array and use InfoGrab to collect host information. After logs are sent back to R&D engineers, the R&D engineers check for risks in the host configuration.
Method 2: Perform a manual check by referring to the following document.
3.3 Why Should I Add Hosts During HyperMetro Inspection?
The check on host configuration during HyperMetro inspection aims to:
1. Ensure that the ALUA configuration of each host takes effect if third-party multipathing software is used.
2. Ensure that the PDL parameter settings are correct for VMware ESX hosts.
This check is not required if Huawei UltraPath is used and the host is not a VMware ESX host.
3.4 How Do I Confirm the Multipathing Software Type?
If the following commands are executed successfully and the command output is displayed, UltraPath is used.
l Windows/AIX/Solaris: upadm show vlun type=hypermetro
l Linux: upadmin show vlun type=hypermetro
l VMware: esxcli upadm show vlun -t hypermetro
The HP-UX operating system does not support UltraPath and thereby can use only third-party multipathing software.
If the host is a FusionSphere host, you can also view the multipathing type of the host on a web page. On the Computing page, select a host and click Configure Stor... on the right.
l If Huawei is selected, Huawei UltraPath is used.
l If Universal is selected, third-party multipathing software is used.
3.5 What Should I Do If the Permission Denied Message Is Displayed During FusionSphere Host Inspection?
You can use only the gandalf account to add a FusionSphere host. However, this account does not have the root permission. If you do not select Switch to user root, the inspection fails due to insufficient permission and the "Permission Denied" message is displayed.
Solution: Add the FusionSphere host using the gandalf account, select Switch to user root, and perform inspection again.
For versions earlier than V2R5C00RC4, the FusionSphere host type is unavailable. You can choose the Linux host type instead.
3.6 What Should I Do If a Storage Array Cannot Be Added for Inspection (Login to DeviceManager Is Successful)?
Different protocols are used for logins to DeviceManager and CLI. Use an SSH tool, such as PuTTY, to check whether a login to CLI is successful. If the login fails, the customer's network may be restricted, so an SSH-based login is not allowed. This requires the confirmation of the customer.
If the SSH-based login is normal, contact Chen Yuanhui/WX418880 for help.
3.7 How Do I Configure the Initiators in H3C Cloud OS?
The underlying layer of H3C Cloud OS is CentOS, and the kernel of CentOS is the same as that of Red Hat. Therefore, for details about the HyperMetro ALUA configuration of H3C Cloud OS, see the configuration method for Red Hat 7.x in the HyperMetro Configuration Guide for Huawei SAN Storage Using OS Native Multipathing Software.
H3C Cloud OS is different from H3C CAS. Do not configure the ALUA configuration of the H3C Cloud OS using the method for H3C CAS.
3.8 How Do I Configure the Initiators in H3C CAS?
Only the following two H3C CAS versions support HyperMetro storage configuration: E0303 and E0306.
1. Configure the initiators of the CAS host on the storage array.
Server OS | Storage Array Configuration | |||||
Storage | OS Setting | Third-Party Multipathing Software | Switchover Mode | Special Mode Type | Path | |
H3C CAS 0303 | Local storage array | Linux | Enabled | Special mode | Mode 0 | Optimal path |
Remote storage array | Linux | Enabled | Special mode | Mode 0 | Optimal path/Non-optimal path | |
H3C CAS 0306 | Local storage array | Linux | Enabled | Common ALUA | N/A | Optimal path |
Remote storage array | Linux | Enabled | Common ALUA | N/A | Optimal path/Non-optimal path | |
2. Configure the multipathing software on the host.
When the special mode is used, add the following contents to the configuration file (/etc/multipath.conf) of the multipathing software on the host:
devices {
device {
vendor "HUAWEI"
product "XSG1"
path_grouping_policy group_by_prio
getuid_callout "/lib/udev/scsi_id --whitelisted --device=/dev/%n"
prio alua
path_checker tur
path_selector "round-robin 0"
failback immediate
fast_io_fail_tmo 15
dev_loss_tmo 30
}
}
When the common ALUA mode is used, add the following contents to the configuration file (/etc/multipath.conf) of the multipathing software on the host:
devices {
device {
vendor "HUAWEI"
product "XSG1"
path_grouping_policy group_by_prio
getuid_callout "/lib/udev/scsi_id --whitelisted --device=/dev/%n"
prio alua
path_checker tur
path_selector "round-robin 0"
failback immediate
fast_io_fail_tmo 5
dev_loss_tmo 30
}
}
3. Configure the HBA.
− If the operating system is H3C CAS 0306 and HBA model is Emulex 16002, contact H3C engineers to update the HBA driver version to 0:11.0.0.0.
− If the HBA model is QLogic QLE2672, ensure that the HBA driver version is 8.07.00.18-k.
4. Check and confirm the multipathing configuration.
After the configuration, the path information is displayed, as shown in the following figure.
3.9 How Do I Enable SSH for VMware ESX?
1. Log in to vSphere vCenter and select the host for which the SSH service needs to be enabled. Then, choose Configuration > Security Profile > Services > Services Properties. In the dialog box that is displayed, select SSH and click Options.
2. In the SSH (TSM-SSH) Options dialog box, select Start and stop manually and click Start.