Safe Mode

A node enters safe mode when there are unrecoverable errors. Any node in a cluster can enter this mode. The purpose of this mode is to detect system configuration failures early and avoid future failures, such as system crashes.

Examples of unrecoverable errors are when there are inconsistencies between the system and the running configuration or when the cluster configuration did not merge properly with the existing configuration.

As part of merge error recovery, nodes joining a cluster are automatically restarted so the merge error can be fixed. If the restart cannot correct the merge error, the node will enter safe mode.

Another example is that a TA Series node could enter safe mode when unlicensed cluster ports are used in an offline configured map. (It is recommended to use only licensed ports in map configurations.)

A node will automatically enter safe mode.

When a node is in safe mode:

■

The node displays a banner indicating it is in safe mode. (Refer to Cluster Safe and Limited Modes.)

■

An SNMP trap is sent to notify the user when the mode changes.

■

Configured traffic continues to be forwarded.

■

Traffic provisioning is not allowed on the affected node. Any other configuration remains as is.

■

If the standby node in the cluster is in safe mode, it can still become the leader leader in clustering node relationship (formerly master) if the current leader fails or switches over, but the database on the standby node may not be in sync, so it is not recommended to continue in that state. Instead, take immediate action to recover the node.

■

In safe mode, the member node follower in clustering node relationship (formerly slave or non-master)s in the cluster do not process any incoming traffic configuration from the cluster leader.

■

SSH is not supported in safe mode if Enhanced Cryptography is enabled. This is a requirement of FIPS. If you are working in an Enhanced Cryptography environment and the node enters in to SAFE mode, you will get a 'connection refused' error. Restart the device to recover the node from SAFE mode. SSH will work again.

When a node is in safe mode and you try do any operations that are not allowed in safe mode, the UI displays the message shown in Cluster Safe and Limited Modes.

When safe mode has been detected, collect information and report it to Gigamon Technical Support. Refer to Collect Information for Technical Support. To recover from safe mode, reload the node.