US 12,474,943 B2
Machine learning based predictions of upgrade testing outcomes for information technology environments
Parminder Singh Sethi, Ludhiana (IN); Shelesh Chopra, Bangalore (IN); and Kanika Kapish, Muzaffarnagar (IN)
Assigned to DELL PRODUCTS L.P., Round Rock, TX (US)
Filed by Dell Products L.P., Round Rock, TX (US)
Filed on Jan. 7, 2022, as Appl. No. 17/570,423.
Prior Publication US 2023/0221973 A1, Jul. 13, 2023
Int. Cl. G06F 9/455 (2018.01); G06F 8/65 (2018.01); G06F 8/71 (2018.01)
CPC G06F 9/45558 (2013.01) [G06F 8/65 (2013.01); G06F 8/71 (2013.01); G06F 2009/45562 (2013.01)] 11 Claims
OG exemplary drawing
 
1. A method for predicting upgrade outcomes entailing prospective product upgrades on enterprise information technology (IT) environments, the method comprising:
receiving an upgrade pre-test request from a customer environment, wherein the upgrade pre-test request comprises a product identifier (ID) associated with an enterprise IT product;
identifying an existing virtual machine group mapped to the enterprise IT product, wherein identification of the existing virtual machine group comprises performing a lookup in a virtual machine group database using the product ID;
obtaining virtual machine feature information for the existing virtual machine group, wherein the virtual machine feature information comprises:
an environment configuration specifying a set of hardware settings for computing hardware being emulated by virtual machines of the existing virtual machine group;
a product version installation chain reflecting a history of enterprise IT product versions, wherein the product version installation chain is specified in a product database of an analysis farm; and
a set of observed upgrade issues that manifested following an installation of any of the enterprise IT product versions, wherein the observed upgrade issues comprise:
a lack of permissions to access one or more directories;
a lack of storage space to accommodate an upgrade installation;
thermal rise across hardware components supporting the existing virtual machine group; and
upgrade incompatibilities;
applying a machine learning algorithm to the virtual machine feature information to generate a high-dimensional analytic subspace, wherein the machine learning algorithm is cluster analysis;
identifying, within the high-dimensional analytic subspace, a plurality of upgrade prediction groups, wherein the plurality of upgrade prediction groups comprise a first upgrade prediction group associated with a successful installation and a second upgrade prediction group associated with a failed installation;
projecting, for the customer environment, customer environment feature information onto the high-dimensional analytic subspace to select an upgrade prediction group of the plurality of upgrade prediction groups, wherein the upgrade prediction group of the plurality of upgrade prediction groups is the first upgrade prediction group;
inferring an upgrade recommendation for the enterprise IT product, based on the upgrade prediction group, wherein the upgrade recommendation is to proceed with an upgrade associated with an upgrade pre-test request; and
providing the upgrade recommendation to the customer environment.