Microsoft Technology Licensing, LLC

Rahul Nigam

Nidhi Verma

Mark Raymond Gilbert

Parminder Pal Singh

Chandramouleeswaran Krishnaswamy

Peter Kenneth Harwood

Mangalam Rathinasabapathy

Jianfeng Cai

Mustafaraj M Dhrolia

Brian G O'Connor

Dmitry Valentinovich Kholodkov

Esha Sharma

Neetha Sumana Tuluri

Sudharsan Ganesan

Vinod Mukundan Menon

Andrei Nicolae

Rohan Khanna

Zohar Raz

Divyachapan Sridharan Padur

Daniel Oh

Jay Lu

Priyanka Gundeli

Eladio Guzman

Hans Christian Andersen

Ankur Jauhari

Roberta Cannerozzi

Donovan Allen Isaak

Pritvinath Obla

Min Shao

Jeremy Haubold

Shobana Krishnamoorthy

Henry Chen

Srigopal Chitrapu

Amber Litteken

Sandhya Shahdeo

Bhuvnesh Kumar

Mark R Gilbert

Vinod Menon

Systems and methods for deploying dependent updates include mechanisms for requiring that dependency information be provided for each update so that dependent updates may be identified. Update dependencies are tracked so that dependent updates are not deployed until parent updates have been completed. Deployment sequencing is implemented on top of existing asynchronous deployment policies so that asynchronous workflow remains intact and unaltered. The asynchronous workflow is upgraded to a synchronous (i.e., sequential) workflow for updates having dependencies to ensure that updates having dependencies are applied in the correct order.

Deployment sequencing for dependent updates

A data processing system implements obtaining a set of input parameters associated with an update to be deployed to a plurality of server farms of a cloud-based service, wherein each server farm includes a primary replica configured to handle user traffic and a disaster recovery replica configured to handle user traffic responsive to a failure of the primary replica; determining temperature information for each of the server farms, ranking the server farms based on the temperature information to determine an order in which an update is to be deployed to the server farms; iteratively deploying the updates to the primary replicas of the server farms according to the ranking until an deployment threshold has been satisfied; and iteratively deploying the updates to the primary replicas of server farms for which the primary replicas have not yet been updated and to the disaster recovery replicas of the server farms.

Techniques for deploying changes to improve reliability of a cloud service

A logging system for an orchestration system that implements a rollout service to deploy updates to a cloud service includes: an orchestrator service host computer hosting the rollout service; a service bus connecting the orchestrator service host computer with a network on which the cloud service is provided; and a database to which the rollout service records a log of a deployment of an update, the log comprising an entry for each of a number of subsets of network components supporting the cloud service. Each record in the log refers to a payload version deployed to a corresponding subset of the network components supporting the cloud service. In response to detection of an issue with the deployment of the update in a particular subset of the network components, the rollout service retains a record in the log of each payload version deployed by changing a payload version number to be negative in each record of the log prior to a restart of the deployment back to an earlier subset of the network components with a new payload version to avoid the issue.

Negative numbering to log web service update attempts

The techniques disclosed herein enable systems to safely deploy a plurality of upgrade variants to different resource units that provide a service by utilizing small-scale deployment and validation. To deploy upgrade variants, a system receives a selection of upgrade variants from a feature group and automatically selects an appropriate set of resource units at which to deploy the upgrade variants. The system is further configured to collect and analyze telemetry data from the set of resource units to determine if any problems have occurred as a result of the deployed upgrade variants. By analyzing the telemetry data, the system can also identify one or more upgrade variants that are causing the problems. In response, the system can remove the identified variants and proceed with deployment of the remaining upgrade variants.

Dark deployment of infrastructure cloud service components for improved safety

A method of and system for implementing a safe deployment of a first feature in a computing environment is carried out by receiving registration information for the first feature, the registration information including a second feature on which the first feature depends or a feature type on which the first feature depends. Based on the second feature or the feature type, a target build version that meets dependency requirements of the second feature or the feature type is then identified before obtaining a list of build versions for a plurality of computing devices in the computing environment. From among the plurality of computing devices, one or more computing devices that have build versions that meet the target build version are then identified before marking the identified computing devices as being available for deployment of the feature and advancing deployment of the feature to those computing devices.

Method and system for implementing safe deployment of features

A data processing system implements managing the deployment of updates to a cloud-based service by deploying an update to one or more components of a cloud-based service according to a deployment plan. The deployment plan defines a plurality of stages in which the update is deployed to a subset of the components associated with a different subset of users of a userbase. The system implements receiving signal data that includes information regarding the performance of the update at each stage of the deployment plan; analyzing the signal data to determine whether one or more trigger conditions of halt and recovery rules associated with the update have been satisfied; and automatically halting deployment of the update to the one or more components of the cloud-based service responsive to at least one of the trigger conditions associated with the halt and recovery rules associated with the update having been satisfied.

Automatically halting cloud service deployments based on telemetry and alert data

A data processing system implements scalable, secure, and reliable targeted release (TR) deployments of updates in cloud-based service. The TR deployment framework is provided that solves the technical problem of facilitating deployment of updates to targeted release customers. TR customers are associated with pre-deployment requirements (PDRs) associated with accreditation and/or validation tasks that must be completed before the update may be deployed to the customer base of the TR customers. A subset of the TR customer base is provided with segregated access to complete these accreditation tasks and/or validation tasks before the update is provided to the remainder of the userbase of the TR customer. This approach ensures that the industry standards and/or customer requirements are met before the update is deployed to the entire userbase of the TR customer.

Targeted release updates for cloud service deployments

A data processing system implements obtaining policy information for an update to be deployed to one or more components of a cloud-based service identifying a plurality of rings of the cloud-based service to which the update is to be deployed; analyzing the policy information to determine configuration information for a plurality of installer instances each associated with one or more rings or one or more stages of a ring to which the update is to be deployed; obtaining payload information for each respective installer instance, the payload information indicating one or more payloads associated with the update to be deployed to the respective one or more rings or one or more stages of the ring associated with the respective installer instance; and deploying the one or more payloads for each of the one or more rings or the one or more stages of the ring associated with each installer instance.

Partitioned deployment of updates to cloud service based on centerally updated configuration store

A data processing system is implemented for detecting changes to infrastructure components, and extracting metadata associated with the changes. The data processing system also implements grouping the changes based on the metadata, ranking the groups of changes based on past incidents of service outages, and displaying the ranked groups of changes to a user.

Ranking changes to infrastructure components based on past service outages

Disclosed herein is a system for improving the user experience in the face of a regression by returning resources that offer a service to a 'last known good' upgrade. In other words, the state of the resources is reconfigured to scale back from recent upgrade(s), the deployments of which likely caused the regression, to a previous upgrade that is known to have little or no effect on the user experience. To identify a problem, the system collects performance data from different resource units that make up a cloud-based platform. The performance data is collected for each upgrade event in a sequence of upgrade events that are currently deployed or being deployed. The system continually tracks and analyzes qualification data collected for each of the deployed upgrade events. The system can tag an upgrade event as the last known good upgrade event when the collected qualification data satisfies predefined qualifications.

Tagging a last known good upgrade event for automatic rollback based on detected regression

A data processing system implements obtaining a set of first input parameters associated with a first update to be deployed to one or more components of a cloud-based service; providing the set of first input parameters to a machine learning model to obtain a first deployment policy for the first update; analyzing the set of first input parameters using the machine learning model to generate the first deployment policy, the machine learning model being trained to analyze input parameters associated with an update to be deployed to the cloud-based service and to generate a deployment policy for the update, the deployment policy identifying a set of rings for deploying the update and when the update is to be deployed to a subset of the userbase of the cloud-based service associated with that ring; and executing the first deployment policy to deploy the update to the one or more components.

Dynamic ring structure for deployment policies for improved reliability of cloud service

In non-limiting examples of the present disclosure, systems, methods and devices for detecting and classifying service issues associated with a cloud-based service are presented. Operational event data for a plurality of operations associated with the cloud-based application service may be monitored. A statistical-based unsupervised machine learning model may be applied to the operational event data. A subset of the operational event data may be tagged as potentially being associated with a code regression, wherein the subset comprises a time series of operational event data. A neural network may be applied to the time series of operational event data, and the time series of operational event data may be flagged for follow-up if the neural network classifies the time series as relating to a positive code regression category.

Automated detection of code regressions from time-series data

Systems and methods for automatically detecting and mitigating errors in a cloud computing environment. One example method includes receiving, from a telemetry server, telemetry data for the cloud computing environment, detecting an error within the cloud computing environment based on the telemetry data, determining an error type for the error based on the telemetry data, determining an impact severity for the error based on the telemetry data, and when the error type is a reportable error type and the impact severity exceeds a predetermined threshold, performing a mitigation action.

Automated detection and classification of dynamic service outages

Systems and methods for automatically recovering from software regression in a cloud computing environment. One example method includes determining, with an electronic processor, that a batch software update has been applied to the cloud computing environment. The method includes, in response to determining that a batch software update has been applied, transmitting a problem request to an event listener server. The method includes receiving, from the event listener server, a problem statement including a stack trace. The method includes determining, based on the stack trace, a software feature indicator. The method includes transmitting the software feature indicator to a root cause analyzer. The method includes receiving, from the root cause analyzer, a change list indicator and a relevancy score associated with the change list indicator. The method includes performing a mitigation action based on the change list indicator when the relevancy score exceeds a relevancy threshold.

Software regression recovery via automated detection of problem change lists

Systems and methods are described for verifying functionality of software. A set of code that is to be validated is identified. A first configuration is determined for the set of code that configures the code as a first build for validation. The first build is released for a first validation process. Prior to completion of validation of the first build, a second configuration is determined for the set of code that configures the code as a second build for validation. The second build is released for a second validation process prior to completion of validation of the first build. The first and second validation process are staged so that the first and second builds can be reverted independently of one another in the event of a validation issue. The first and second validation process are independently completed in the absence of a validation issue.

Growing community of inventors

Bothell, WA, United States of America

Rahul Nigam