| CPC G06F 9/48 (2013.01) [G06F 9/46 (2013.01); G06F 9/4881 (2013.01); G06F 9/5083 (2013.01); G06F 11/0709 (2013.01); G06F 11/0772 (2013.01); G06F 11/2028 (2013.01); G06F 11/3006 (2013.01); G06F 11/3409 (2013.01)] | 20 Claims |
|
1. A method comprising:
generating a plurality of device trees for respective ones of a plurality of devices in a datacenter, the plurality of device trees respectively comprising a plurality of nodes corresponding to a device, one or more components of the device and one or more alerts associated with the one or more components, the plurality of device trees including at least a first device tree generated for a first device in the datacenter, the first device tree comprising at a first level a root node representing the first device, at a second level a plurality of component nodes each coupled to the root node and each representing a different component of the first device, and at a third level a plurality of alert nodes each coupled to a corresponding one of the component nodes and each representing a different type of alert generated for the component represented by the corresponding component node, the plurality of device trees further including a second device tree generated for a second device in the datacenter, the second device being different than the first device, the second device tree having a root node, component nodes and alert nodes at respective first, second and third levels of the second device tree;
assigning weights to the plurality of nodes of the plurality of device trees, wherein assigning weights comprises, for a given one of the first and second device trees generated for the respective first and second devices, assigning a weight of a first set of weights to the root node, assigning weights of a second set of weights to respective component nodes at the second level of the given device tree, and assigning weights of a third set of weights to respective alert nodes at the third level of the given device tree;
computing rates of degradation for the respective ones of the plurality of devices based, at least in part, on the assigned weights, wherein for the given device tree, a rate of degradation is computed for the corresponding first or second device based at least in part on total weights of respective ones of the component nodes, wherein the total weight of a particular one of the component nodes is computed at least in part as a function of the weight assigned from the second set of weights to the particular component node and the weights assigned from the third set of weights to respective ones of the alert nodes coupled to the particular component node;
predicting workloads for the respective ones of the plurality of devices using one or more machine learning models;
generating a ranking of the respective ones of the plurality of devices based, at least in part, on at least one of the rates of degradation and the predicted workloads; and
managing the devices based at least in part on the generated ranking;
wherein the steps of the method are executed by a processing device operatively coupled to a memory.
|