Maximizing Data Security and Efficiency with CLAM in NetApp

CLAM, it stands for the Continuous, Liveness, Availability, Monitor. This is a kernel mechanism that basically helps to verify that all the communication paths between the nodes are accessible and also can be able to perform the CTRAN [CLAM Transaction].

CLAM determines which node can do the following:

1. Server the data requests.

2. Participate in the CLAM.

The CLAM uses the heartbeat mechanism and if there's is no response from the node then the node is marked or considered as OOQ [Out of Quorum] node. A node is declared in quorum if it is responding to the ping request sent by the master node. By default, it sends 13 consecutive pings if the node doesn't respond back then it is OOQ node.

How CLAM Works?

1. When the node lost connectivity, it initiates the planned takeover.

2. If RDB in NetApp on any minority node is offline, then no changes in data will happen and only read requests are accepted.

3. The Partition event occurs when the node goes from the minority to majority and vice versa. After all that CLAM checks other node by heartbeat. If receive pings, then In Quorum otherwise OOQ.

Now, how to enable the CLAM in NetApp.

cluster kernel-service config modify -node "node name" -kcs-enable-takeover "True/False"

Now, there's a question arises that why CLAM in NetApp introduced.

To understand this, we need to understand the before ONTAP 9, NetApp's condition.

Scenario before ONTAP 9 CLAM --- Suppose there's a four-node cluster and out of which two nodes are OOQ then whole cluster service goes down. Even the rest two clusters are good and can serve the data.

Scenario after ONTAP 9 CLAM --- CLAM evaluates quorum and considers node failover state. Now instead of 0 node alive it will consider 2 nodes to serve the data that results in increase in cluster uptime.