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


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


Host Using Virtual Storage Resources of an FC SAN Device Fails to Safely Restart

Publication Date:  2014-05-15 Views:  110 Downloads:  0

Issue Description

An FC SAN storage device were used in the system, and the storage virtualization function was enabled.
  •  The task for restarting a host in the storage virtualization cluster suspended.
  •  VMs running on other hosts in the same cluster cannot be started, and the cause of the failure was that volume attaching timed out.
  •  The VM HA task suspended when the progress reached 70%. 

Alarm Information


Handling Process

You can use either of the following solutions to rectify the fault:
  •  In the scenario where FC SAN storage device is used and the storage virtualization function is enabled, do not safely stop or power off a host, or run the reboot command to restart a host. If you do need to power off or restart a host, forcibly power off or restart it in the BMC system.
  •  Upgrade FusionCompute to V100R003C10SPC500 or later.

Root Cause

1. To restart a host, the hypervisor performs operations in the following sequence: stop management and service processes > stop VMs running on the hosts one by one > disassociate data stores > stop network and storage services > detach file systems > stop the host.
2. Based on the preceding processing logics, the VND process will be first stopped before the device for which memory swapping fails is detached. Therefore, the data store fails to be detached.
3. In the scenario where the FC SAN storage device is used, communication on the management plane is interrupted after the network service is stopped. Then the hypervisor handles the storage processing logics and attempts to send message to other nodes in the VIMS cluster. However, communication on the management plane is interrupted, and the nodes cannot receive messages from the hypervisor. As a result, the host restart task suspends.