Description

When the RAID card of a node in the HCI cluster is abnormal (for example, the card runs slowly), the storage performance of the cluster will be degraded or even cause the production/storage to fail. The RAID status anomaly detection function can detect the abnormality of the RAID card, notify the administrator in time, and isolate the physical node corresponding to the RAID card when the card is stuck to avoid affecting the stability of the entire cluster and interrupting storage as the whole or cluster production.

Precautions

1. The node must configure IPMI. Otherwise, the node cannot be recovered remotely after being isolated.
2. Alert prompts might fails if the RAID card has failed.
3. In a two-node Virtual Storage cluster (in the scenario where no witness node is configured), the virtual machine may experience split-brain when one of the nodes has isolated.

Prerequisites

None

Steps

1. Login to the HCI WebUI console and navigate to System > Advanced. Under Node, enable the RAID Card Status Check.

2. When the RAID card is at fault, an alert will be prompted: RAID card status of node (xxx.xxx.xxx.xxx) is abnormal. Virtual Machines on the node will automatically HA or failover to another node to avoid cluster production interruption.