Node Creation with Unknown MAC address(es)#

A reset or powercycle of a node triggers a DHCP client request that embeds the node's MAC address. A head node with an interface that is listening on that private cluster network and recognizes the MAC address responds with an IP address that is associated with that MAC, unless directed to ignore that node. A ClusterWareAI ™ head node can be so directed to ignore the known-MAC node by using a _no_boot attribute (see _no_boot).

By default, a ClusterWareAI DHCP server ignores the incoming DHCP client request if it does not recognize the incoming MAC address. Ignored nodes are visible on the Unknown Nodes tab of the Nodes Page in the ClusterWareAI GUI. To override this default, either enable the Accept unknown nodes? switch on the Cluster Settings page of the ClusterWareAI GUI or run the following command:

cw-clusterctl --set-accept-nodes True

If a cluster is set to accept unknown nodes, any head node that shares the same database can add that new MAC to the shared ClusterWareAI database, assign to it the next available node index and associated IP address, and attempt to boot the node.

When all new nodes with previously unknown MAC addresses are merged into the ClusterWareAI cluster, return to the default functionality by disabling the Accept unknown nodes? switch in the ClusterWareAI GUI or by running the following command:

cw-clusterctl --set-accept-nodes False

If multiple new nodes concurrently initiate their DHCP client requests, then the likely result is a jumbled assignment of indices and IP addresses. Cluster administrators often prefer nodes in a rack to have ordered indices and IP addresses.

Ordered assignment of previously unknown nodes can be accomplished by:

  • Leaving the default setting of accept unknown nodes as false and adding nodes manually using the Unknown Nodes tab in the ClusterWareAI GUI. See Create Compute Node for details.

  • Setting accept unknown nodes to true and performing the initial node resets or powercycling for each individual node in sequence. Allow each node to boot and get added to the database before initiating the next node's DHCP request.

  • Setting accept unknown nodes to true and allowing the nodes to concurrently initiate their DHCP client requests. After all nodes are added to the ClusterWareAI software, perform subsequent cw-nodectl update actions and reboot the updated nodes. For example:

    cw-nodectl -i n10 update index=100
cw-nodectl -i n11 update index=101
cw-nodectl -i n12 update index=102
cw-nodectl -i n10,n11,n12 reboot