receiving input/output (I/O) requests of a workload at a file server of a node coupled to persistent storage media, wherein different types of I/O requests have corresponding latency bounds;

processing the I/O requests of the workload according to an expected I/O latency using a pool of threads executing on one or more processors of the node;

maintaining a number of the threads as active threads of the pool to run on the processors;

measuring an I/O latency for storing data of the I/O requests on the persistent storage media;

and

in response to the measured I/O latency for storing the data exceeding a predetermined threshold, increasing the number of active threads from the pool deployed to increase processing of the I/O requests to run to completion on the active threads based on the corresponding expected I/O latency bounds for the I/O request type.