RAN Edge Site Upgrade via Boot Disk Replacement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing solutions for upgrading cell sites in a RAN are laborious and time-consuming, requiring multiple waves and long maintenance windows, leading to significant system downtime and revenue loss.
Innovation Solution
A method for rapidly upgrading multiple machines across cell sites by downloading an upgraded boot disk, preserving data, and using local agents to manage successful or failed upgrades, with a control plane server overseeing the process and reverting changes if necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional upgrade methods are used for cell sites, then system reliability is maintained through careful sequential updates, but upgrade time increases significantly requiring multiple waves and long maintenance windows
Solution Approach 1:
The system segments the upgrade process into distinct phases: preparation phase where the upgraded boot disk is downloaded and validated, execution phase where multiple VMs are upgraded in parallel, and rollback phase if needed. This segmentation allows concurrent operations while maintaining system stability.
Solution Approach 2:
The system performs preliminary actions by downloading and validating the upgraded boot disk before the maintenance window begins. Local agents prepare upgrade packages in advance, and data snapshots are created beforehand, enabling rapid execution during the actual upgrade window without time-consuming preparation steps.
2Reliability
If multiple upgrade waves are performed to upgrade 6000+ cell sites, then system stability is maintained, but revenue loss increases due to extended downtime
Solution Approach 1:
The system merges multiple sequential upgrade waves into a single parallel execution wave. Local agents at different cell sites execute upgrades simultaneously, combining what would have been 48 separate waves into one coordinated operation, thereby reducing total downtime from 24-72 hours to a single maintenance window.
Solution Approach 2:
The system implements feedback mechanisms where local agents report upgrade status to a control plane server in real-time. The server monitors progress and can trigger coordinated rollbacks if issues arise, maintaining system stability while enabling parallel execution. This feedback loop replaces the conservative sequential approach with controlled parallelism.
3Loss of time
If a single maintenance window is used to upgrade all cell sites in parallel, then upgrade time is reduced, but system complexity increases due to coordination requirements
Solution Approach 1:
The control plane server acts as an intermediary between the upgrade management system and local agents at cell sites. It coordinates the parallel upgrades by distributing upgraded boot disks, monitoring execution status, and managing rollback procedures, thereby simplifying the complexity of coordinating thousands of simultaneous upgrades across distributed locations.
Solution Approach 2:
Local agents at each cell site perform self-service upgrades by automatically downloading boot disks, creating data snapshots, executing the upgrade process, and reporting status without manual intervention. This automation reduces the coordination complexity that would otherwise require extensive human management of parallel upgrades.
4Reliability
If data is preserved during upgrade process, then reliability is improved through rollback capability, but storage requirements increase
Solution Approach 1:
The system implements discard and recover by creating temporary snapshots of data during the upgrade process for potential rollback, then discarding these snapshots after successful upgrade verification. This approach maintains reliability through rollback capability while minimizing long-term storage requirements, as the preserved data is only held temporarily during the maintenance window.
Data Source
AI summary
Some embodiments of the invention provide, for a RAN (radio access network), a method of rapidly upgrading multiple machines distributed across multiple cell sites, each particular machine of the multiple machines executing one or more base station applications. The method downloads a second boot disk for each of the multiple machines at each of the multiple cell sites, the second boot disk including an upgraded version of a first boot disk currently used by each of the multiple machines. For each particular machine, the method (1) powers off the particular machine, (2) creates a copy of data stored by a data disk of the particular machine to preserve data stored currently on the data disk, (3) replaces the first boot disk of the particular machine with the second boot disk that is the upgraded version of the first boot disk, and (4) powers on the particular machine.