Before
Network engineers likely reconciled management-console versions with the inventory, looked up the approved version for each model, and checked maintenance freezes manually. They then grouped devices into waves, created campaign changes, updated compliance records, and linked each device to the appropriate change.
How it works
- Collect observed and approved firmware versions and stop without writes when baselines are missing or management-system coverage is below 50%.
- Classify each device, update or create compliance records, correct inventory discrepancies, and identify upgrade gaps.
- Group gaps into freeze-aware maintenance waves, request the campaign and wave changes, link compliance records, and return a consolidated summary.
Workflow details
- Evidence collection: An on-demand run, documented for Monday at 06:30, reads network devices, load balancers, approved firmware baselines, existing compliance records, management-system observations, team assignments, and change-freeze periods.
- Validation and classification: The workflow requires an approved baseline and checks that each management source reports at least half of the expected devices before writing anything. It classifies equipment as compliant, needing an upgrade, on an out-of-reference branch, requiring an inventory correction, unknown to the inventory, or without a reference.
- Record maintenance: It creates or updates one compliance record per device, corrects inventory versions when observed data differs, and records devices seen by management systems but missing from the inventory.
- Wave planning: Upgrade gaps are grouped by model, responsible team, and maintenance window, with store work planned by region at night and data-center work on Sundays. Windows are moved ahead of detected freezes or flagged when an exception is required.
- Change preparation: The workflow requests a parent campaign change and one child change per upgrade wave, includes the affected equipment and implementation details, links compliance records to the relevant change, and returns counts, change numbers, and write-failure warnings.
What changed
Each run produces a traceable compliance state and an actionable upgrade plan from current evidence. Coverage and baseline checks happen before records are changed, while each planned wave receives its device scope, owner, timing, freeze context, change request, and linked compliance records.
Work removed
Manually comparing device versions with baselines, checking management-system coverage, correcting inventory records, creating compliance entries, grouping devices into maintenance waves, checking change freezes, creating related changes, and linking compliance records.
Expected outcome
A repeatable weekly firmware review with grouped, traceable upgrade planning; the documented demo scenario expects seven maintenance waves instead of an unstructured fleet-wide follow-up.
