Network Interface Backup (NIB)

EtherChannel Backup is a variant of EtherChannel that is used for high-availability only. EtherChannel Backup allows an aggregated adapter to have a backup. If all adapters that compose the aggregation fail, then communication is switched to the backup adapter until any adapter in the main channel recovers. A variant of it is Network Interface Backup (NIB).

In the NIB mode of operation, there is only 1 adapter in the main channel and a backup adapter. While NIB by itself does not provide better bandwidth than the physical adapter, it can be used to work around switch failures. Usually port aggregation requires all adapters to be connected to the same switch, which makes the switch the single point of failure. By using NIB, and by connecting the primary and backup adapters to different switches, communication will not be lost by the failure of a single switch.

To help detect loss of network reachability (in addition to detecting failures in the adapter and its connection to the switch), NIB allows specifying an address to be pinged. If the given address cannot be reached after a given number of attempts (both specified when NIB is defined), then the current active adapter is considered down, resulting in the backup adapter taking over communication. Setting reasonable values for the Number of Retries option is important to ensure smooth operation of NIB: if the value is not enough to cover the period during which the switch is reconfiguring itself, it is likely that there will be multiple (false) takeover operations until one of the adapters becomes the owner of the aggregation. Such extra takeover activity makes real (or desired) takeover operations take much longer than intended.

As an initial guideline, setting Number of Retries to 10 should correct the false takeover problem in cases where communication outages are around 8 seconds.

The false takeover scenario can be identified by examining the AIX® error log. In case the scenario occurs, entries like the following may appear:
  • ECH_PING_FAIL_PRMRY
  • ECH_PING_FAIL_BCKP
  • GXENT_LINK_DOWN

When Number of Retries is set to an adequate value, then error log entry ECH_CHAN_FAIL may be the only one to be generated.

Since NIB uses a single adapter as primary, an EtherChannel-enabled switch is not required.