About FMC High Availability
To ensure the continuity of operations, the high availability feature allows you to designate redundant FMCs to manage devices. The FMCs support Active/Standby high availability where one appliance is the active unit and manages devices. The standby unit does not actively manage devices. The active unit writes configuration data into a data store and replicates data for both units, using synchronization where necessary to share some information with the standby unit.
Active/Standby high availability lets you configure a secondary FMC to take over the functionality of a primary FMC if the primary fails. When the primary FMC fails, you must promote the secondary FMC to become the active unit.
Event data streams from managed devices to both FMCs in the high availability pair. If one FMC fails, you can monitor your network without interruption using the other FMC.
Note that FMCs configured as a high availability pair do not need to be on the same trusted management network, nor do they have to be in the same geographic location.
Caution |
Because the system restricts some functionality to the active FMC, if that appliance fails, you must promote the standby FMC to active. |
Note |
Triggering a switchover on FMC immediately after a successful change deployment can lead to preview configuration not working on the new active FMC. This does not impact policy deploy functionality. It is recommended to trigger a switchover on the FMC after the necessary sync is completed. Similarly, when FMC HA synchronization is in degraded state, triggering a switchover or changing roles could make FMC HA to damage the database and it can become catastrophic. We recommend that you immediately contact Cisco Technical Assistance Center (TAC) for further assistance to resolve this issue. This HA synchronization can end up in degraded state due to various reasons. The Replacing FMCs in a High Availability Pair section in this chapter covers some of the failure scenarios and the subsequent procedure to fix the issue. If the reason or scenario of degraded state matches to the scenarios explained, follow the steps to fix the issue. For other reasons, we recommend that you contact TAC. |
About Remote Access VPN High Availability
If the primary device has Remote Access VPN configuration with an identity certificate enrolled using a CertEnrollment object, the secondary device must have an identity certificate enrolled using the same CertEnrollment object. The CertEnrollment object can have different values for the primary and secondary devices due to device-specific overrides. The limitation is only to have the same CertEnrollment object enrolled on the two devices before the high availability formation.
SNMP Behavior in FMC High Availability
In an SNMP-configured HA pair, when you deploy an alert policy, the primary FMC sends the SNMP traps. When the primary FMC fails, the secondary FMC which becomes the active unit, sends the SNMP traps without the need for any additional configuration.
Roles v. Status in FMC High Availability
Primary/Secondary Roles
When setting up Firepower Management Centers in a high availability pair, you configure one Firepower Management Center to be primary and the other as secondary. During configuration, the primary unit's policies are synchronized to the secondary unit. After this synchronization, the primary Firepower Management Center becomes the active peer, while the secondary Firepower Management Center becomes the standby peer, and the two units act as a single appliance for managed device and policy configuration.
Active/Standby Status
The main differences between the two Firepower Management Centers in a high availability pair are related to which peer is active and which peer is standby. The active Firepower Management Center remains fully functional, where you can manage devices and policies. On the standby Firepower Management Center, functionality is hidden; you cannot make any configuration changes.
Event Processing on FMC High Availability Pairs
Since both FMCs in a high availability pair receive events from managed devices, the management IP addresses for the appliances are not shared. This means that you do not need to intervene to ensure continuous processing of events if one of the FMC fails.
AMP Cloud Connections and Malware Information
Although they share file policies and related configurations, FMCs in a high availability pair share neither Cisco AMP cloud connections nor malware dispositions. To ensure continuity of operations, and to ensure that detected files’ malware dispositions are the same on both FMCs, both primary and secondary FMCs must have access to the AMP cloud.
URL Filtering and Security Intelligence
URL filtering and Security Intelligence configurations and information are synchronized between Firepower Management Centers in a high availability deployment. However, only the primary Firepower Management Center downloads URL category and reputation data for updates to Security Intelligence feeds.
If the primary Firepower Management Center fails, not only must you make sure that the secondary Firepower Management Center can access the internet to update threat intelligence data, but you must also use the web interface on the secondary Firepower Management Center to promote it to active.
User Data Processing During FMC Failover
If the primary FMC fails, the Secondary FMC propagates to managed devices user-to-IP mappings from the TS Agent identity source; and propagates SGT mappings from the ISE/ISE-PIC identity source. Users not yet seen by identity sources are identified as Unknown.
After the downtime, the Unknown users are re identified and processed according to the rules in your identity policy.
Configuration Management on FMC High Availability Pairs
In a high availability deployment, only the active FMC can manage devices and apply policies. Both FMCs remain in a state of continuous synchronization.
If the active FMC fails, the high availability pair enters a degraded state until you manually promote the standby appliance to the active state. Once the promotion is complete, the appliances leave maintenance mode.
FMC High Availability Disaster Recovery
In case of a disaster recovery situation, a manual switchover must be performed. When the primary FMC - FMC1 fails, access the web interface of the secondary FMC - FMC2 and switch peers. This is applicable conversely also in case the secondary (FMC2) fails. For more information, see Switching Peers in the FMC High Availability Pair.
For restoring a failed FMC, refer to Replacing FMCs in a High Availability Pair.
Single Sign-On and High Availability Pairs
FMCs in a high availability configuration can support Single Sign-On, but you must keep the following considerations in mind:
-
SSO configuration is not synchronized between the members of the high availability pair; you must configure SSO separately on each member of the pair.
-
Both FMCs in a high availability pair must use the same identity provider (IdP) for SSO. You must configure a service provider application at the IdP for each FMC configured for SSO.
-
In a high availability pair of FMCs where both are configured to support SSO, before a user can use SSO to access the secondary FMC for the first time, that user must first use SSO to log into the primary FMC at least once.
-
When configuring SSO for FMCs in a high availability pair:
-
If you configure SSO on the primary FMC, you are not required to configure SSO on the secondary FMC.
-
If you configure SSO on the secondary FMC, you are required to configure SSO on the primary FMC as well. (This is because SSO users must log in to the primary FMC at least once before logging into the secondary FMC.)
-
FMC High Availability Behavior During a Backup
When you perform a Backup on a FMC high availability pair, the Backup operation pauses synchronization between the peers. During this operation, you may continue using the active FMC, but not the standby peer.
After Backup is completed, synchronization resumes, which briefly disables processes on the active peer. During this pause, the High Availability page briefly displays a holding page until all processes resume.
FMC High Availability Split-Brain
If the active FMC in a high-availability pair goes down (due to power issues, network/connectivity issues), you can promote the standby FMC to an active state. When the original active peer comes up, both peers can assume they are active. This state is defined as 'split-brain'. When this situation occurs, the system prompts you to choose an active appliance, which demotes the other appliance to standby.
If the active FMC goes down (or disconnects due to a network failure), you may either break high availability or switch roles. The standby FMC enters a degraded state.
Note |
Whichever appliance you use as the secondary loses all of its device registrations and policy configurations when you resolve split-brain. For example, you would lose modifications to any policies that existed on the secondary but not on the primary. If the FMC is in a high availability split-brain scenario where both appliances are active, and you register managed devices and deploy policies before you resolve split-brain, you must export any policies and unregister any managed devices from the intended standby FMC before re-establishing high availability. You may then register the managed devices and import the policies to the intended active FMC. |
Upgrading FMCs in a High Availability Pair
Cisco electronically distributes several different types of updates periodically. These include major and minor upgrades to the system software. You may need to install these updates on FMCs in a high availability setup.
Warning |
Make sure that there is at least one operational FMC during an upgrade. |
Before you begin
Read the release notes or advisory text that accompanies the upgrade. The release notes provide important information, including supported platforms, compatibility, prerequisites, warnings, and specific installation and uninstallation instructions.
Procedure
Step 1 |
Access the web interface of the active FMC and pause data synchronization; see Pausing Communication Between Paired FMCs. |
Step 2 |
Upgrade the standby FMC. |
Step 3 |
Upgrade the other FMC. |
Step 4 |
Decide which FMC you want to use as the standby. Any additional devices or policies added to the standby after pausing synchronization are not synced to the active FMC. Unregister only those additional devices and export any configurations you want to preserve. When you choose a new active FMC, the FMC you designate as secondary will lose device registrations and deployed policy configurations, which are not synced. |
Step 5 |
Resolve split-brain by choosing the new active FMC which has all the latest required configurations for policies and devices. |
Troubleshooting FMC High Availability
This section lists troubleshooting information for some common FMC high availability operation errors.
Error |
Description |
Solution |
||
---|---|---|---|---|
You must reset your password on the active FMC before you can log in to the standby. |
You attempted to log into the standby FMC when a force password reset is enabled for your account. |
As the database is read-only for a standby FMC, reset the password on the login page of the active FMC. |
||
500 Internal |
May appear when attempting to access the web interface while performing critical FMC high availability operations, including switching peer roles or pausing and resuming synchronization. |
Wait until the operation completes before using the web interface. |
||
System processes are starting, please wait Also, the web interface does not respond. |
May appear when the FMC reboots (manually or while recovering from a power down) during a high availability or data synchronization operation. |
|
||
Device Registration Status:Host <string> is not reachable |
During the initial configuration of a FTD, if the FMC IP address and NAT ID are specified, the Host field can be left blank. However, in an HA environment with both the FMCs behind a NAT, this error occurs when you add the FTD on the secondary FMC. |
|
||
Device Registration Status:Host <string> is not reachable |
The error occurs when adding FTD device to the secondary FMC center in a high-availability deployment where both the secondary FMC and the FTD device are behind NAT. |
On the standby FMC web interface, click . Under the pending device registration table, click the IP address of the pending device, and then change the IP address to the public IP address of the FTD. OR
|