Hypershield - Release notesπ
Release notes for the latest Hypershield versions are listed here:
Hypershield v10.0.0.0 (2026-08-21) release notesπ
Synopsisπ
Cisco Hypershield 10.0.0.0 is the initial generally available release of the Cisco Hypershield offering.
This Hypershield release applies to the following components:
Hypershield Controller v10.0.0.0
Hypershield Agent Gateway (AGW) v10.0.0.0
Hypershield Network Enforcer (FWA) v10.0.0.0
Hypershield Deployment Appliance v10.0.0.0
Cisco Cloud Control (August 2026 or newer)
New featuresπ
Software featuresπ
Component affected |
Feature name |
Description |
|---|---|---|
Hypershield Controller |
Management User interface: Timescape |
Provides browser-based management of policies, objects, deployments, events, and system status. |
Cisco Cloud Control |
Management User interface: Cisco Cloud Control |
Integration and access to the Management User Interface through Ciscoβs cloud-based management platform as an option. |
Hypershield Controller |
Programmatic Access: RestAPI |
Automates supported network-object and security-policy operations through versioned REST endpoints. |
Hypershield Controller |
Leader Controller |
On-premise virtual appliance that hosts management services and maintains the authoritative policy configuration and propagates it down to all followers. |
Hypershield Controller |
Follower Controller |
An on-premise virtual appliance that hosts the control plane services and distributes approved policy to connected enforcement points and reports deployment status. Multiple Follower Controllers may be deployed inside or across multiple data centers connected to the same Leader Controller. |
Hypershield Controller |
Backup/Restore |
Protects and restores the controller state, configuration, policy data, and authorization to meet recovery requirements. |
Hypershield Controller |
Authorization/RBAC |
Integration with existing OIDC Providers for authorization and control of administrative access through assigned roles and permissions. |
Hypershield Controller |
Policy Management |
Supports creating, reviewing, deploying, and monitoring security policies based on Network Object Groups using Labels and/or VLAN/VRF/CIDR and Port and Protocol constructs. |
Hypershield Controller |
Network Object Groups |
Organizes reusable network objects into small groups that are then labeled or directly leveraged for policy management. |
Hypershield Controller |
Telemetry |
Centralized configuration of logging groups and destinations for operational and flow data via Syslog and IPFIX. |
Hypershield Deployment Appliance |
Hypershield Controller Deployment Automation |
A management VM (virtual machine) that automates controller deployment and initial leader and follower cluster configuration leveraging the hypervisor management API. |
Hardware featuresπ
Component affected |
Feature name |
Description |
|---|---|---|
Hypershield Agent Gateway (AGW) |
Automated Policy Placement |
Delivers approved policies to the appropriate Hypershield Network Enforcer inside the Smart Switch just in time based on the NX-OS Switch Configuration. |
Hypershield Agent Gateway (AGW) |
Logging Export to Multiple Destination |
Streams audit and enforcement events to multiple configured syslog destinations directly from the Hypershield Agent Gateway running in each Smart Switch. |
Hypershield Network Enforcer (FWA) |
IPFIX Telemetry Export to Multiple Locations |
Streams IPFIX flow records to multiple configured collectors destinations directly from the Hypershield Network Enforcers running in each Smart Switch. |
Hypershield Network Enforcer (FWA) |
Stateful L3/L4 Policy Enforcement |
Enforces Layer 3 and Layer 4 policy on unicast network flows steered to the Network Enforcer in each Smart Switch. |
Hypershield Network Enforcer (FWA) |
Active/Active State Synchronization |
Synchronizes connection state between paired Smart Switches for resilient traffic processing. |
Caveats and limitationsπ
Information related to current limitations of the platform. Best practices are provided to help guide the Operator.
Description |
Component |
Impact |
Workaround |
|---|---|---|---|
Policy and network object payload size |
Hypershield Controller |
A SmartSwitchNetworkPolicy or NetworkObjectGroup serialized above 1.5 MB can fail when the follower Kubernetes API materializes it. |
Keep each object below 1.5 MB. Split large policies or object groups into smaller reusable objects. |
OIDC refresh-token flow |
Hypershield Controller |
Users must authenticate again when the ID token expires, even when the identity provider supplies offline access. |
Plan for periodic re-authentication and choose an appropriate ID-token lifetime in the identity provider. |
Event-history CSV export |
Hypershield Controller |
A CSV export contains only the events displayed on the current page, not the complete filtered result set. |
Export each page and combine the files, or review the complete history in Timescape. |
Leader enrollment event visibility |
Hypershield Controller |
Leader enrollment events are not shown in event history, so enrollment does not create a visible audit entry in the UI. |
Record leader enrollment through the organization change-control process. |
Telemetry configuration replacement |
Hypershield Controller / Hypershield Agent Gateway |
A pre-existing statically named syslog or IPFIX ConfigMap can prevent a new UI-managed configuration from reaching previously configured Smart Switches. |
Remove the obsolete ConfigMap from the affected follower cluster, then reapply the telemetry configuration and verify delivery. |
VM live migration |
Hypershield Network Enforcer |
Live migration of a protected workload VM resets established TCP sessions that traverse DPU stateful inspection. |
Drain connections or schedule migration during a low-traffic maintenance window. |
Policy updates during control-plane disconnection |
Hypershield Agent Gateway |
A disconnected Smart Switch continues enforcing its last installed policy in a fail-closed posture, but it cannot receive new policy. |
Restore follower-to-switch connectivity before deploying policy changes. |
Policy restore after Smart Switch power cycle |
Hypershield Network Enforcer / Agent Gateway |
A power cycle clears in-memory DPU policy. Traffic is fail-closed until the switch reconnects and the control plane restores the current policy. |
Use a correctly configured HA partner to maintain enforcement while the rebooted switch recovers. |
Broadcast, unknown-unicast, and multicast traffic |
Hypershield Network Enforcer |
BUM traffic bypasses the DPU and is not evaluated by Hypershield policy. |
Use NX-OS controls such as storm control or IGMP snooping; do not rely on Hypershield policy for BUM filtering. |
Bidirectional policy on L2-only switches |
Hypershield Network Enforcer |
On a switch without routed interfaces, policy can be unidirectional if both source and destination VLANs are not steered to the DPU. |
Configure both VLAN IDs in the service firewall block and validate traffic in both directions. |
Inter-VRF mode counters |
Hypershield Network Enforcer |
Inter-VRF mode is enabled by default. The first packet makes two NPU-to-DPU passes, which can double-count dropped packets on the service Ethernet interface. |
Account for the double count in monitoring. Disable inter-VRF mode only for an intra-VRF-only design and only through an approved procedure. |
Very large policy deployment |
Hypershield Network Enforcer / Agent Gateway |
Applying more than 500,000 rules can temporarily reduce throughput while the receiving Smart Switch programs the policy. |
Schedule very large deployments during lower traffic and monitor switch health and policy status until synchronization completes. |
Policy references a missing VRF |
Hypershield Controller |
The policy remains Pending and rules that depend on the missing VRF are not enforced on that Smart Switch. |
Configure the referenced VRF. If its absence is intentional, treat Pending as expected for that target. |
Policy references a missing VLAN |
Hypershield Controller |
The policy remains Pending and rules that depend on the missing VLAN are not enforced on that Smart Switch. |
Configure the referenced VLAN. If its absence is intentional, treat Pending as expected for that target. |
Mixed IPv4/IPv6 endpoint rule |
Hypershield Controller |
A policy that pairs principal and resource from different IP families remains Pending and is not enforced. |
Use IPv4-to-IPv4 or IPv6-to-IPv6 source and destination pairs. |
Policies reference mismatched VLAN IDs in Network Object Groups |
Hypershield Controller |
Policies using VLAN-based objects are only enforced when principal and resource VLANs match. Use VRF-based objects for inter-VLAN traffic. |
For inter-VLAN enforcement policies, create policies using network object groups that leverage VRF, not VLAN, IDs. |
Open issuesπ
Platform issues that are actively tracked.
Description |
Component |
Impact |
Workaround |
|---|---|---|---|
Policy Deployment status shows in-sync when the deployment is still pending |
Hypershield Controller |
After a policy is edited or redeployed, the deployment status may continue to show βIn Syncβ instead of βPendingβ until the updated version reaches the network switches. This window is more pronounced if |
Wait for policy update to refresh. |
Smart Switch HA failover |
Hypershield Network Enforcer |
Failover can take up to 10 seconds, and some connections can be interrupted during convergence. |
Design applications to retry transient failures and validate the HA peer path before production use. |
Policy status updates can be lost |
Hypershield Controller |
In some cases, policy status updates are lost when the switch becomes disconnected from the control plane. This can result in the UI showing a policy as Pending even though it has been successfully applied on the switch. |
Users can make a change to the policy (e.g., add a new rule) and reapply it to force a status update. Alternatively, users can check the switch logs or use CLI commands to verify the policy status on the switch. |
Policy status updates stop being sent after a switch reboot |
Hypershield Network Enforcer |
After a Smart Switch reboots, it may stop sending policy status updates to the Hypershield Controller. This can result in the UI showing a policy as Pending even though it has been successfully applied on the switch. |
Users can make a change to the policy (e.g., add a new rule) and reapply it to force a status update. Alternatively, users can check the switch logs or use CLI commands to verify the policy status on the switch. |
Feedback