RAC-VOTINGDISK

Voting

1. What is a Voting Disk?

It acts as a tibrakers during communication failover, consistent hard-beat information from all nodes is sent to the voting disk.

The Voting Disk contains a list of all nodes that are currently members of the cluster, as well as information about the health and status of each node. When a node joins or leaves the cluster, the Voting Disk is updated with the new membership information.

2. How Voting Disks Work? During normal operation, each node in the cluster sends heartbeats to the other nodes, indicating that it is still operational. If a node fails to send a heartbeat for a specified period of time, the other nodes in the cluster will assume that the node has failed and will remove it from the cluster membership.

3. When do we need to take backup of the Voting Disk? After Installation After Node add/delete After perform voting disk add delete

4. How to perform backup of Voting Disk? dd if=/location of voting disk of=/backup location

5. What happens if I lose my voting disk(s)?

Losing a Voting Disk in an Oracle RAC environment can lead to cluster node evictions and potentially disrupt the cluster's availability and stability. It's crucial to have redundancy and backups for Voting Disks to avoid such issues.

6. Potential impacts of losing the Voting Disk(s)?

Split-Brain Scenario: If all of the nodes in the cluster lose connectivity to the Voting Disk(s), each node may assume that it is the only active node in the cluster. This can cause a split-brain scenario, where each node tries to take control of the resources that were running on the other nodes. This can lead to data corruption and service interruption.

Inability to Manage the Cluster: Without the Voting Disk(s), the Clusterware may not be able to manage the cluster effectively. For example, if a node fails, the Clusterware may not be able to relocate resources to another node, as it cannot determine which node is the correct owner of the resources.

Loss of Membership Information: The Voting Disk(s) contain information about the membership of the cluster, including the list of nodes and their health status. If the Voting Disk(s) are lost, this information will be lost as well. This can make it difficult to determine the status of the cluster and troubleshoot problems.

7. How to prevent from losing the Voting Disk(s)?

To mitigate the risk of losing the Voting Disk(s), it is recommended to configure the Voting Disk(s) in a highly available configuration, such as using mirroring or using multiple Voting Disks. This can help to ensure that the Voting Disk(s) remain available even if one or more disks fail. It is also recommended to periodically backup the Voting Disk(s) to a secure location, in case they need to be restored in the event of a failure.

6. How much I/O activity should the voting disk have? Approximately 2 read + 1 write per second per node.

7. Why does Oracle Clusterware use an additional 'heartbeat' via the voting disk? Oracle uses this implementation because Oracle clusters always have access to a shared disk environment.

Activity

1. Recover the Oracle voting disk when its backup is missing

We can restore the voting disk from the OCR backup

Step 1: Disable the CRS automatic start service

Step 2: Restart the node

Step 3: Verify that the CSR service did not start

Step 4: Clear the header of the voting disk

Step 5: Start the cluster, Start the ASM with a PFILE

Step 7: Create a disk group, Create a SPFILE and restart the ASM

Step 9: Restore the current OCR backup and Replace the voting disk

Step 11: Enable the CRS automatic start service

Step 11: Enable the CRS automatic start service

Step 12: Cross-check the OCR disk status,Check the voting disk details,Check the ASM disk groups

2. Restore loss of all VOTE disks

Stop CRS on all the nodes

Start CRS in exclusive mode only

Create New Diskgroup

Restore/Move/Replace Votedisk

Stop CRS on Node 1

Start CRS on both nodes

Check Cluster Status