Sliding Timer Expiry to Report Faults to Ethernet Ring Protection Mechanism

Abstract

The G.8032 Ethernet ring protection (ERP) mechanism detects the failure in a network connected in a ring and will automatically switch traffic path to ensure that uninterrupted traffic provided even in case of a node or path failure. To avoid different race conditions and unnecessary switching operations, it uses different timers. One of such static timers configured today is the hold-off timer. After detecting a failure, ERPS will not switch immediately, and it will wait for this time; if the failure is not cleared before the timer expiry, then actual switching happens. This waiting period is added to avoid unnecessary switching due to link flaps. So even when an actual failure is detected, ERPS will wait till this timer expiry, until then traffic loss will be seen. Currently, the timer range is between 0 and 10 s which is configured by the operator. The default value of this timer is 0 s, and in this case, the timer will be disabled. It is best advised to configure and enable the timer to avoid false reporting, frequent switching, and flapping. But, if the timer is configured very less, frequent unnecessary switching will happen and if the timer is kept very high, traffic loss will be seen for that wait time. The objective of this work is to propose a sliding/adaptive timer expiry algorithm incorporated inside the ERPS module itself instead of a static hold-off timer.