The “split brain” scenario is an exceptional circumstance where two healthy nodes in an HA pair both believe themselves to be the Active node, and that the other node is the Standby.
This scenario represents an unhealthy HA pair, and must be resolved.
The “split brain” scenario can occur after a failed forced failover operation. Specifically:
1.The healthy Standby node becomes an Active node.
2.The unhealthy Active node fails to become the Standby node.
3.The unhealthy Active node is repaired. Both nodes are now healthy and Active, and each also believes the other node in the HA pair to be the Standby node. This is the “split brain” scenario.
To resolve the “split brain” scenario, perform a Forced Standby operation from the repaired Active node. This forces the repaired Active node to become the Standby node in the HA pair.
原文链接：Recovering from a Split Brain Scenario
The manual translation is as follows:
Split Brain Scenario occurs in an HA pair (High Availability pair) (distributed system). Both healthy nodes consider themselves as active nodes and the other node is a standby node. Special case.
This situation means that this is an unhealthy HA pair. If you want the system to operate normally, you must solve this problem.
Split Brain Scenario will occur after a failed forced failover (failover) (network abnormality or active node failure). details as follows:
- The original active node fails, and the normally operating backup node becomes the active node.
- A failed active node cannot be converted to a backup node.
- When the failed active node is repaired, there are two active active nodes in the system (both consider themselves as Master nodes), and the two nodes consider each other as backup nodes. This is the Split Brain Scenario.
In order to solve the Split Brain Scenario, a forced backup operation needs to be run on the repaired active node. This will force the repaired active node to become a backup node, thereby solving the Split Brain Scenario.
An HA pair is two storage systems(nodes) whose controllers are connected to each other directly.In this configuration,one node can take over its partner’s storage to provide continued data service if the partner goes down.
You can configure the HA pair so that each node in the pair shares access to a common set of storage,subnets and tape drives,or each node can own its own distinct set of storage.
The controller are connected to each other through an HA interconnection.This allows one node to serve data that resides on the disks of its failed partner node.Each node continually monitors its partner,mirroring the data for each other’s nonvolatile memory(NVRAM or NVMEN).The interconnect is internal and requires no external cabling if both controllers are in the same chassis.
Takeover is the process in which a node takes over the storage of its partner.Giveback is the process in which that storage is returned to the partner.Both processes can be initiated manually or configured for automatic initiation.
The manual translation is as follows:
an HA pair is two storage systems (nodes) directly connected by a controller (controller). In this configuration, if one node fails, another node can take over the storage of its partner node and provide continuous data services (high availability).
You can manually configure an HA pair so that each node in the pair shares a common set of storage, subnet, and tape drive access, or each node can have its own different storage set.
The controllers communicate with each other through an HA interconnect. This allows a node to provide services to the data on the disk of its failed partner node. Each node continuously monitors its partners, mirroring data for each other's non-volatile memory (NVRAM or NVMEN). The HA interconnection is internal, and if the two controllers are in the same chassis, no external wiring is required.
Takeover is the process by which a node takes over the storage of its partner. Giveback is the process of returning storage to the partner. Both processes can be started manually or configured to start automatically.