CliDeck
PlatformLog in
Workspace

Arista EOS Zero Touch Provisioning Cleanup After Reset: Preventing Old Automation From Reappearing

ITAD & Refurbishment
#Arista#EOS#Zero Touch Provisioning#ZTP#ITAD#ITAD & Refurbishment#itad#network-device-processing
← Back to Blog

Short answer

Use this as an owner-authorized workflow for arista eos zero touch provisioning cleanup after reset: preventing old automation from reappearing. Capture the console session, use the approved device-specific procedure, verify the post-reset or post-recovery state, and attach evidence-style results to the asset record. This article does not describe certified data destruction.

Arista EOS Zero Touch Provisioning Cleanup After Reset: Preventing Old Automation From Reappearing

A reset that removes the startup configuration can still leave an EOS switch ready to enter Zero Touch Provisioning on the next boot. For an authorized resale or redeployment, explicitly control ZTP, then verify after reload that old automation will not reappear on the bench network.

Scope and ownership boundary

Use this workflow only for an owned or customer-authorized asset that has an approved ITAD, return, refurbish, or redeployment work order. The owner defines whether the destination switch should be ZTP-ready for a new deployment or intentionally kept out of provisioning for resale. The technician must never attach a wiped device to an unknown provisioning network and must not disable or alter automation outside the approved asset scope. The asset owner approves the disposition and recovery scope; the technician captures evidence, follows the exact-model decision gate, and stops when an observation does not match the work order. This is not a process for bypassing an active organization, recovering someone else’s service, or retaining customer configuration.

Preflight checklist

  1. Match the physical asset to the approved work order and keep serial numbers, license records, management identifiers, and account information only in the private record.
  2. Isolate the device from production networks. Use a known-safe bench network or local console that contains no customer credentials, provisioning service, or production automation.
  3. Capture a private before-state: model family, software or boot version, current role, observed management state, planned action, operator, and timestamp.
  4. Confirm that the approved action matches the device’s model, software family, and boot context. A command or menu option from a different release is a stop condition, not a substitute.
  5. Prepare a private pass/fail form before changing state. The result must show what was observed, not just that a command was entered.

Model and version gate

The observed DCS-7050QX-32 used EOS 4.14.9M and displayed a ZTP message when no startup configuration was present. EOS behavior and available configuration files can vary by release and boot state. Confirm the exact software version, the current startup configuration condition, and the owner-approved ZTP disposition before executing a persistent change.

Safe execution checklist

  1. Collect a private version and boot baseline while the switch is isolated from DHCP, provisioning, and production management services.
  2. Decide with the owner whether the end state should be an intentionally disabled ZTP posture or a controlled clean device for a new authorized automation domain.
  3. When the approved goal is to prevent old automation, use the EOS ZTP disable action and allow its device-controlled reload to finish.
  4. After boot, review the private console behavior and relevant flash configuration artifacts for the expected post-reset state.
  5. Keep the switch disconnected from unknown network services until the runbook’s clean-state and ownership checks pass.

Commands and controlled observations

Run only the commands that are valid for the exact device and approved context. Do not paste production addresses, image locations, account data, or credentials into the public work record.

show version
zerotouch disable
dir flash:

On the observed EOS family, the ZTP disable action triggers a reload. Treat that reload as part of the change window. The directory review is private evidence; do not publish filenames, provisioning content, addresses, or automation artifacts from the asset.

Pass/fail signals

Pass: After the approved reload, the switch reaches the expected console state without beginning unwanted provisioning, and the private review agrees with the owner-approved ZTP posture.

Fail or hold: Any automatic provisioning attempt, unexpected retained configuration, unrecognized flash artifact, unsupported command behavior, or disagreement about whether ZTP should remain available. Hold the switch and escalate before another reset.

A successful command alone is not sufficient. The technician must observe the specified boot or management result, keep the device isolated until review is complete, and escalate any mismatch to the asset owner.

Sanitized field note

Sanitized field note: A DCS-7050QX-32 running EOS 4.14.9M displayed the Zero Touch Provisioning message after boot without a startup configuration. The observed zerotouch disable action initiated a reload, and the private flash review included a ZTP-related configuration artifact.

The note is a sanitized operational observation, not a replacement for the exact model and release check. It contains no customer names, hostnames, addresses, MAC addresses, credentials, license identifiers, file paths, or raw console transcript.

Evidence artifacts for the private record

  • Authorized work-order reference and disposition decision.
  • Model, software or boot version, and recovery context observed before the action.
  • Exact reset, recovery, boot, or cleanup path selected, including any confirmation result.
  • Boot completion result, clean-state or management-state evidence, and explicit pass/fail outcome.
  • Exception owner and reason when the device is held rather than released.
  • Technician identity, timestamp, and the sanitized handoff summary.

Never place raw console output, addresses, configuration exports, account names, customer topology, certificates, keys, passwords, licensing details, or recovery-file locations in a public resale listing.

Make the handoff repeatable in CliDeck

Create a CliDeck runbook with an ownership confirmation before the destructive or recovery step. Add a required model/version gate, a private evidence form, and a post-boot pass/fail checkpoint. Use a required branch for “ZTP intentionally enabled” versus “ZTP intentionally disabled,” then require a post-reload console observation. This prevents a reset-only completion from masking a provisioning risk. If the device reports an unexpected state, mark the runbook as held and route it to the asset owner instead of continuing toward resale.

For related owner-authorized workflows, see the Arista EOS recovery guide, the Dell ONIE decision guide, and the ExtremeXOS factory-default checklist.

Final disposition gate

Mark the switch ready only when the owner-approved ZTP disposition persists through reload and no prior automation starts on the isolated bench. A reset without this check is not yet a verified resale state.

Safety note: This guidance applies only to devices your team owns or is authorized to process. Use the approved, device-specific procedure, customer policy, and your organization’s sanitization and evidence procedures.

Who this is for

This guide is for ITAD teams, refurbishment teams, network engineers, and operations teams who need to perform an owner-authorized reset, recovery, verification, or documented device-processing workflow on network devices that your team owns or is authorized to process.

When to use this

Use this guide when:

  • Your team owns or is authorized to process the device.
  • You need to reset, zeroize, recover, reimage, verify, or document a network device.
  • The result must be attached to an asset, resale, reuse, refurbishment, or recycling workflow.

When not to use this

Do not use this guide when:

  • You do not own or have authorization to work on the device.
  • The device is outside your approved change, recovery, or processing scope.
  • The task may expose customer data, secrets, licenses, or credentials that you are not allowed to view or store.
  • You need a certified data destruction procedure and your organization has not provided one.

Workflow

  • Confirm ownership, processing scope, device identity, and approved procedure.
  • Connect through the console or approved management path and capture the session context.
  • Run the reset, zeroize, recovery, reimage, or verification steps required by the article and vendor documentation.
  • Reboot or reconnect as needed and verify the expected clean, default, or recovered state.
  • Attach logs, verification output, exceptions, and final pass/fail/exception status to the asset record.

Verification checklist

  • Device identity matches the asset record.
  • Console or session log is attached to the asset.
  • The reset, zeroize, recovery, or reimage method is recorded.
  • The post-reboot state matches the expected clean or recovered state.
  • Verification commands and outputs are captured.
  • Exceptions or failures are quarantined instead of marked ready.

Resale and evidence checklist

Before marking the device ready for resale, capture:

  • Device model and serial number
  • Console transcript or session log
  • Reset, zeroize, reimage, or recovery method used
  • Any confirmation prompts and operator decisions
  • Post-reboot state
  • Visible default or clean-state prompt
  • Verification commands and outputs
  • Hardware alarms or boot errors
  • License, entitlement, or customer identifiers handled according to policy
  • Exceptions, failures, or quarantine decisions
  • Final pass/fail/exception status attached to the asset record

What not to publish

Do not publish raw transcripts that include customer names, public IP addresses, VPN settings, license keys, entitlement IDs, serial numbers that must remain private, passwords, tokens, certificates, or other sensitive identifiers. Keep sensitive evidence in the customer-approved internal system and publish only sanitized examples.

Common mistakes

  • Processing a device without owner authorization or the correct asset record.
  • Treating a factory reset as certified data destruction.
  • Failing to capture confirmation prompts, reboot state, verification output, or exceptions.
  • Publishing raw logs that contain customer identifiers, licenses, keys, passwords, or configuration remnants.

CliDeck workflow fit

CliDeck can help turn this procedure into a repeatable workflow. Console Server Go is useful for portable rack-side console access and shared remote support. Net Controller is better for ITAD, refurbishment, and batch processing where multiple devices need barcode tracking, approved scripts, logs, API updates, and evidence-style results.

Net Controller can help operators connect devices, scan assets, run approved workflows, capture console logs, require verification steps, and attach the result to the asset record.

FAQ

Is this a certified data destruction procedure?

No. This article describes a practical, owner-authorized workflow and verification approach. Certification depends on your organization’s process, evidence requirements, customer contract, and any formal certification program that applies.

Does CliDeck provide vendor OS images?

No. CliDeck does not distribute vendor operating system images. Recovery or reimage workflows can use images already present on the device or images provided by the customer where the customer has appropriate rights and licenses.

Should I publish the full console log?

Usually no. Keep complete evidence in the approved internal system and publish only sanitized examples. Console logs can contain serial numbers, license details, customer names, IP addresses, keys, passwords, or configuration remnants.

Can CliDeck help with this workflow?

Yes. Console Server Go helps with portable rack-side console access and shared remote support. Net Controller is better for batch ITAD, refurbishment, recovery, provisioning, barcode/API tracking, and evidence-style results.

Related guides