EX436 sample questions

Question 1

As a system administrator for NovaTech, you are deploying a new Red Hat Enterprise Linux (RHEL) High Availability cluster. Before creating the cluster, you must authenticate the nodes so that the `pcs` command can communicate with the `pcsd` daemon across all systems. Which command is the correct standard method in RHEL 8/9 to accomplish this?

Accepted Answer

pcs host auth node1 node2 -u hacluster

Answer

pacemaker auth init node1 node2

Answer

corosync-keygen --auth node1 node2

Answer

pcs cluster auth node1 node2 -u hacluster

Question 2

You are automating the deployment of a Pacemaker cluster using the Red Hat Enterprise Linux `ha_cluster` Ansible system role. Which TWO variables must be defined in your playbook to successfully set the cluster name and the authentication password for the `hacluster` user? (Select TWO)

Accepted Answer

ha_cluster_cluster_name

Accepted Answer

ha_cluster_hacluster_password

Answer

ha_cluster_name

Answer

ha_cluster_auth_password

Answer

pacemaker_cluster_name

Question 3

AcmeCorp operates a critical two-node High Availability cluster. The network link between Site A (Node 1) and Site B (Node 2) completely fails, but both nodes remain powered on and functioning.

To prevent a split-brain scenario where both nodes attempt to mount the shared storage concurrently, how does Corosync's votequorum subsystem handle this by default in a properly configured two-node setup without a qdevice?

```mermaid
graph TD
subgraph Site A
Node1[Node 1]
end
subgraph Site B
Node2[Node 2]
end
Node1 -.-x|Network Link Failed| Node2
```

Accepted Answer

The `two_node=1` setting artificially grants quorum to the node that successfully executes a fencing action against the peer.

Answer

Both nodes lose quorum because neither has the strict majority (>50%) of votes required, shutting down cluster resources on both sides.

Answer

The `wait_for_all` flag allows both nodes to independently achieve quorum, relying entirely on STONITH to resolve the conflict.

Answer

The node with the lowest node ID automatically assumes quorum and fences the other node.

Question 4

True or False: In a standard 3-node RHEL HA cluster without a qdevice, if two nodes fail simultaneously, the single remaining node will maintain quorum and continue running resources because it is the 'last man standing'.

Accepted Answer

False

Answer

True

Question 5

You need to completely wipe out an existing Pacemaker/Corosync cluster configuration on a node and return it to a clean state. Select the correct sub-command to complete this operation: `pcs cluster _____`

Accepted Answer

destroy

Answer

purge

Answer

remove

Answer

delete

Question 6

What is the primary architectural purpose of configuring STONITH (Shoot The Other Node In The Head) in a Red Hat High Availability cluster?

Accepted Answer

To guarantee that a malfunctioning node is completely isolated from shared storage to prevent data corruption.

Answer

To elect a new cluster leader when the designated master node goes offline.

Answer

To provide a heartbeat mechanism over a secondary network interface when Corosync fails.

Answer

To restart failed Pacemaker resources automatically on the same node without manual intervention.

Top vendors

Top exams

Free EX436 sample questions

You are automating the deployment of a Pacemaker cluster using the Red Hat Enterprise Linux `ha_cluster` Ansible system role. Which TWO variables must be defined in your playbook to successfully set the cluster name and the authentication password for the `hacluster` user? (Select TWO)

True or False: In a standard 3-node RHEL HA cluster without a qdevice, if two nodes fail simultaneously, the single remaining node will maintain quorum and continue running resources because it is the 'last man standing'.

You need to completely wipe out an existing Pacemaker/Corosync cluster configuration on a node and return it to a clean state. Select the correct sub-command to complete this operation: `pcs cluster _____`

What is the primary architectural purpose of configuring STONITH (Shoot The Other Node In The Head) in a Red Hat High Availability cluster?

6 more free samples are waiting