Methods and systems for detecting anomalies during IO accesses

摘要

An anomaly in a shared input/output (IO) resource that is accessed by a plurality hosts or clients is detected when a host that is not bound by any QoS policy presents large workloads to a shared IO resource that is also accessed by hosts or clients that are governed by QoS policy. The anomaly detection triggers a response from the hosts or clients as a way to protect against the effect of the anomaly. The response is an increase in window sizes. The window sizes of the hosts or clients may be increased to the maximum window size or in proportion to their QoS shares.

主权项

1. A method of detecting that at least one of a plurality of host computers that are accessing a shared IO resource is not abiding by a quality of service (QoS) policy that is followed by a remaining of the plurality of host computers, wherein the QoS policy governs rates at which the remaining of the plurality of host computers access the shared IO resource, the method comprising:
computing, at periodic intervals, aggregate changes for each parameter of a set of QoS related parameters that is maintained by each of the host computers following the QoS policy, wherein a value of each parameter may change while the corresponding host is accessing the shared IO resource; determining whether each of computed aggregate changes exceeds a threshold value for the corresponding parameter, wherein exceeding the threshold value indicates a possibility that a host computer is accessing the shared IO resource without following the QoS policy; and if each of the computed aggregate changes exceeds the corresponding threshold value, storing a value in a manner accessible by each of the host computers following the QoS policy, wherein the value indicates that a host computer accessing the shared IO resource is not following the QoS policy.