Investigate management data synchronization status and underlying connectivity or load. This can be appropriate when a secondary FortiManager reports a large pending module synchronization backlog during peak network usage. It does not directly address the present situation, where the required correction is confirm the peer has current synchronized modules before initiating the approved failover procedure.

Investigate peer network reachability and HA configuration before relying on failover. This can be appropriate when a FortiManager HA secondary has stopped receiving heartbeat packets but the primary remains online. It does not directly address the present situation, where the required correction is confirm the peer has current synchronized modules before initiating the approved failover procedure.

Use the supported `get system ha` diagnostic command on the relevant appliance. This can be appropriate when a new administrator needs a CLI confirmation of FortiManager HA state when the graphical interface cannot load. It does not directly address the present situation, where the required correction is confirm the peer has current synchronized modules before initiating the approved failover procedure.

The replacement serial receives the approved policy package and appears online. That observation is relevant for the separate case a firewall replacement boots successfully, yet FortiManager continues sending tasks to the failed old serial. Here the decisive evidence is disk utilization returns to a supported level without deleting current management databases.

The staged rollout completes with stable management CPU, memory and disk headroom. That observation is relevant for the separate case a central management appliance exhausts resources during a firmware rollout because all managed devices were scheduled at once. Here the decisive evidence is disk utilization returns to a supported level without deleting current management databases.

Collected diagnostics identify the component responsible or justify the controlled restart. That observation is relevant for the separate case an operator sees high memory usage but immediately reboots FortiManager without collecting diagnostic evidence. Here the decisive evidence is disk utilization returns to a supported level without deleting current management databases.

The transferred device appears only in the intended destination ADOM with appropriate management scope. That observation is relevant for the separate case a device needs to move from one customer ADOM to another due to a contract transfer. Here the decisive evidence is the externally authenticated user can operate only within approved delegated privileges.

Managed device operations follow the supported workflow after the mode decision. That observation is relevant for the separate case a branch FortiGate is registered under an ADOM whose device operating mode does not support an expected workflow. Here the decisive evidence is the externally authenticated user can operate only within approved delegated privileges.

Users assigned to one customer cannot read or modify another customer's ADOM records. That observation is relevant for the separate case an MSSP wants different customers' FortiGates to have separate object libraries and administrator privileges. Here the decisive evidence is the externally authenticated user can operate only within approved delegated privileges.

Forced lock removal without preserving changes can lose valid but unsubmitted configuration. That consequence is more specific to an administrator disconnects while holding an ADOM lock and a peer cannot continue urgent work. For this case, the most direct exposure is selecting the wrong collaboration mode can leave required approval steps unenforced.

Skipping approval sessions can violate separation-of-duties requirements. That consequence is more specific to an operator enables workflow mode and immediately installs a policy package without submitting the session for approval. For this case, the most direct exposure is selecting the wrong collaboration mode can leave required approval steps unenforced.

Treating every policy-package lock as an ADOM-wide lock can unnecessarily block unrelated work. That consequence is more specific to a managed device requires a local interface correction, but the policy package is locked by another administrator. For this case, the most direct exposure is selecting the wrong collaboration mode can leave required approval steps unenforced.

Unsupported object attributes can block installation or produce incomplete device configuration. That consequence is more specific to an installation task fails during object-copy processing because the target device rejects an unsupported security profile field. For this case, the most direct exposure is fleet-wide deployment without a representative validation may spread a bad object change.

Random troubleshooting changes expand outage scope without fixing the indicated failure. That consequence is more specific to a FortiManager operator sees an install log error mentioning a missing dependency but attempts to change unrelated interface settings. For this case, the most direct exposure is fleet-wide deployment without a representative validation may spread a bad object change.

A task marked complete without matching device state can conceal unresolved synchronization. That consequence is more specific to an engineer edits the central policy but an installation task shows the same diff repeatedly despite its claimed success. For this case, the most direct exposure is fleet-wide deployment without a representative validation may spread a bad object change.

The run log contains required device diagnostic output and shows no configuration changes. That observation is relevant for the separate case a network engineer needs a controlled one-time diagnostic command across several managed firewalls without saving a permanent policy. Here the decisive evidence is each eligible test device accepts the script without command errors.

The script is committed under the authorized session and competing edits are prevented. That observation is relevant for the separate case an administrator edits a script in workspace mode without first obtaining the required creation or editing lock. Here the decisive evidence is each eligible test device accepts the script without command errors.

The final configuration remains correct after the next supported installation task. That observation is relevant for the separate case a branch script completes but a later policy install appears to reverse a configuration setting. Here the decisive evidence is each eligible test device accepts the script without command errors.

Read the failed copy detail and align the object with the target FortiOS feature support. This can be appropriate when an installation task fails during object-copy processing because the target device rejects an unsupported security profile field. It does not directly address the present situation, where the required correction is identify the named dependency from the failed task and correct only that referenced object.

Inspect per-target task results and isolate only the failed device for repair. This can be appropriate when an installation fails on one branch and the administrator assumes every target in the batch failed. It does not directly address the present situation, where the required correction is identify the named dependency from the failed task and correct only that referenced object.

Run the supported filesystem and database integrity checks under a controlled recovery procedure. This can be appropriate when a sudden power failure occurs during FortiManager database writes and later operations report storage corruption. It does not directly address the present situation, where the required correction is identify the named dependency from the failed task and correct only that referenced object.

Uncontrolled script timing can disrupt production networking or compliance controls. That consequence is more specific to a scheduled script changes a network parameter outside an authorized maintenance window. For this case, the most direct exposure is misunderstanding script execution context can affect unauthorized policy or device scopes.

Script-only changes can be overwritten by the authoritative centrally managed configuration. That consequence is more specific to a branch script completes but a later policy install appears to reverse a configuration setting. For this case, the most direct exposure is misunderstanding script execution context can affect unauthorized policy or device scopes.

Incompatible commands can leave a fleet with mixed or incomplete settings. That consequence is more specific to a FortiManager script succeeds on one test firewall but fails on an older FortiOS version in another site. For this case, the most direct exposure is misunderstanding script execution context can affect unauthorized policy or device scopes.

Eligible devices show the correct reviewed image and complete maintenance with expected versions. That observation is relevant for the separate case an administrator wants to push approved firmware to many FortiGates but did not review each model's compatibility. Here the decisive evidence is the firewall uses its authorized fallback service when the private server is unavailable.

FortiGuard service tests successfully retrieve updates through the approved proxy path. That observation is relevant for the separate case a FortiManager in a restricted site must download FortiGuard updates through an enterprise web proxy. Here the decisive evidence is the firewall uses its authorized fallback service when the private server is unavailable.

Eligible FortiGates retrieve current antivirus and IPS packages through the approved local server. That observation is relevant for the separate case a group of isolated FortiGates should retrieve signature updates through a centrally controlled management network instead of each using the internet. Here the decisive evidence is the firewall uses its authorized fallback service when the private server is unavailable.

The chosen revision remains listed after the automated retention cleanup runs. That observation is relevant for the separate case an ADOM revision was automatically purged even though the audit team needed to preserve it. Here the decisive evidence is required snapshots are preserved and obsolete revisions are pruned according to policy.

The relevant baseline snapshot has traceable creator and change-context information. That observation is relevant for the separate case a SOC team asks who created a policy baseline revision before a high-risk maintenance event. Here the decisive evidence is required snapshots are preserved and obsolete revisions are pruned according to policy.

The intended device interface state is recovered without unrelated ADOM policy object rollback. That observation is relevant for the separate case an engineer attempts to use an ADOM revision to roll back an out-of-band interface change on one firewall. Here the decisive evidence is required snapshots are preserved and obsolete revisions are pruned according to policy.

The task log lists all three intended branches and each reports the new policy revision. That observation is relevant for the separate case a security team installs a new policy on only one of three approved branches because the installation target list is outdated. Here the decisive evidence is controlled test traffic matches the intended new allow rule and its effective object resolution.

Installation preview resolves every dynamic object for the selected device. That observation is relevant for the separate case an operator attempts to deploy a security policy with a dynamic address object that has no mapping for one target. Here the decisive evidence is controlled test traffic matches the intended new allow rule and its effective object resolution.

A new validation run succeeds after the specific object or mapping issue is repaired. That observation is relevant for the separate case a FortiManager installation task fails at its validation stage but the operator retries unchanged settings repeatedly. Here the decisive evidence is controlled test traffic matches the intended new allow rule and its effective object resolution.