Methods and architecture for load-correcting requests for serverless functions

A computing network with infrastructure for running serverless functions, comprising:

• server nodes communicatively coupled with each other;
• each server node comprising one or more processors and one or more non-transitory computer-readable media storing computer-executable instructions that, when executed by the one or more processors, cause the one or more processors to:
• load an identity of an alternate node of the computing network for running a serverless function, the identity of the alternate node to be stored as an identified alternate node;
• receive, from a client node of the computing network, a request to run the serverless function;
• determine that the serverless function is not currently running at the server node or determine that the server node is currently overloaded;
• and
• based on determining that the serverless function is not currently running at the server node or based on determining that the server node is currently overloaded, redirect the request to the identified alternate node while refraining from assessing whether the serverless function is currently running at the identified alternate node.

A server node for a cloud computing network hosting serverless functions, comprising:

• at least a processor and a memory;
• a front-end part of the server node for receiving, from a client device, a request to run a serverless function;
• a backend part of the server node comprising a function engine for creating a working environment and resources for running the serverless function at the server node;
• and
• a referral component in the front-end part of the server node for directing the request to another function engine in the backend part of another server node in any location in the cloud computing network.