Problem and solution when a shared folder is lost on the LUN connected to iSCSI after the Windows host is restarted


You can perform the following operations when a shared folder is lost on the LUN connected to iSCSI after a Windows host is restarted:
1. Method
The shared folder is created by the server service during startup. If the server service starts before the iSCSI initiator service during server startup, the server cannot create file sharing for the iSCSI disk. For temporary workaround, manually start the server service.
2. Solution
Perform the following steps:
a. Change the startup sequence of server services by using the iSCSI initiator server. Enter sc config lanmanserver depend= MSiSCSI on the CLI (note that there is a space after =).
b. Add the disk volume list to the permanent binding list of Microsoft iSCSI initiator. Enter iscsicli Bindpersistentvolumes on the CLI.

Other related questions:
Solution if a message indicating that the host needs to be restarted is displayed repeatedly when the multipathing software is being reinstalled on a Windows host after it is uninstalled
Solution if a message indicating that the host needs to be restarted is displayed repeatedly when the multipathing software is being reinstalled on a Windows host after it is uninstalled: Delete the Restart key value in the registry. The position where the key value of Restart is stored varies with Windows systems: 1. Windows x86 system: HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\Run 2. Windows x64 system: HKEY_LOCAL_MACHINE\SOFTWARE\Wow6432Node\Microsoft\Windows\CurrentVersion\Run 3. Restart the host.

Problem and solution when some LUNs fail to be mapped when LUNs are mapped in batch
After LUN mappings are canceled in batch in a storage system, the LUNs generated on a host still exist. As a result, the storage system returns "Unmapped" with error code 0x05/0x25/0x00 due to I/Os sent using these LUNs. The interface protocol indicates that the error code performs a LUN switchover but the storage system uses a system lock for the LUN switchover. As a result, LUN mapping fails because no system lock can be obtained. To solve this problem, map the LUNs again. Then check whether the number of LUNs that are successfully mapped is correct.

Method used if multiple hosts share a LUN
You can refer to the following case when multiple hosts share a LUN: To improve the reliability and availability of the system, sometimes multiple hosts must share one LUN. When one LUN is shared by multiple hosts, if a file is used by multiple hosts at the same time, the file may be damaged. To avoid such damages, you must include multiple hosts into a cluster system and manage the hosts. The cluster system provides highly reliable services for customers, having the following features: A cluster is a group of independent servers that function and are managed as a single system. A cluster contains at least two servers with shared data space. Application data of servers is stored in the shared space while their operating systems and program files are saved in their local storage space. Each cluster node is a server and communicates with each other. If a node server is faulty, anther node server will take over its service. Customers are connected to new application servers without feeling service interruption. Servers within a cluster perform backups for each other, which achieves high availability. The following is the most common mutual backup modes: Hot backup. In this mode, servers work in active/standby mode. When applications are running on the active server, the standby server is in the idle state and monitors the running status of the active server in real time. Upon the failure of the active system, the standby server immediately takes over services. Mutual backup. In this mode, both of the servers are running properly without a division of active and standby servers. The two servers run different applications and monitor each other in real time. If one of the two servers is faulty, the other server immediately takes over its applications without service interruption. Duplex. In this mode, two or more servers are running properly and run same applications. The high performance of the cluster is ensured with load balancing and mutual backup. After a cluster is created on the application server, perform the following operations in the storage system: 1. Log in to the ISM and create a host group. For details, refer to section "Creating a Host Group" of the Configuration Guide in the provided documentation CD-ROM. 2. Create hosts in the host group interface. The number and operating system type of hosts must be the same as those of actual hosts. For details, refer to section "Creating a Host" of the Configuration Guide. 3. Map a LUN to the host group so that the LUN can be shared by hosts in the host group. For details, refer to section "Adding a Mapping" of the Configuration Guide. 4. Set up connections between the host and the storage system. For details, refer to section "Connect the Application Server to the Storage Device" of the Configuration Guide. 5. Configure an initiator and connect hosts created on the storage system to the actual physical hosts. For details, refer to section "Configuring an Initiator" of the Configuration Guide. 6. Scan for the host and mount the mapped LUN to the host. For details, refer to section "Making Storage Space Available on an Application Server" of the configuration Guide. Through the above operations, multiple hosts can share one LUN.

How do I solve the DHCP configuration loss problem after device restart

After the device restarts, there is a low probability that the DHCP configuration is lost.

Auto-Config is enabled by default. About two and a half minutes after the device restarts, the Auto-Config module checks whether the configuration file exists and determine whether the Auto-Config process is executed. If the save command is used at this time, files conflict. As a result, the Auto-Config module determines that the configuration file does not exist, and incorrectly executes the Auto-Config process. In this case, information about option fields or intermediate file required for executing the Auto-Config process is not configured on the DHCP server. Consequently, the Auto-Config process is suspended, and the DHCP configuration (including the configuration during configuration recovery and the Auto-Config process) is deleted.

After the device restarts, run the undo autoconfig enable command in the system view to disable Auto-Config.

Solution if the host fails to detect LUNs
Solution if the host fails to detect LUNs: Determine the connection mode at the front end. If the Fibre Channel switch is connected, check whether the switch mode has been configured. If the front end uses the Fibre Channel switch, check with the customer that the switch has been correctly configured with zones and assign the storage system and the host to the same zone. If the port WWN is bonded during zone division, check whether the port binding is correct. Alternatively, use optical fibers to directly connect the storage system to the switch and re-scan for LUNs.

If you have more questions, you can seek help from following ways:
To iKnow To Live Chat
Scroll to top