Scaling for split-networking datapath

A non-transitory machine readable medium storing a program for execution by a set of processing units, the program comprising sets of instructions for:

• providing a plurality of nodes, the plurality of nodes comprising a first node and a second node;
• executing a first set of pods by the first node, the first set of pods comprising a first L4 pod;
• processing a first data message for a first logical router by executing a run-to-completion thread by the first L4 pod;
• assigning a first core of the first node for processing data traffic between a logical network and an external network;
• determining that the first node has reached a predetermined threshold limit in response to a resource request;
• and
• adding a third node to the plurality of nodes for providing resources associated with the resource request;
• and
• scheduling a second pod for the plurality of nodes after adding the third node to the plurality of nodes;
• wherein the first plurality of nodes is associated with a Kubernetes cluster for implementing an edge device.