When you configure a PP protection over ring of 3 or more equipments and one equipment is a virtual NE wich has diferent class of cross-conections, switching the services over virtual NE not working properly. The cross-conections over virtual NE are of high order (Vc4) and over the others equipments, the cross-conections are of low order (Vc12).
LP_RDI ; TU_AIS
When we check the path of the work of services affected, some were working for the fiber that was open and the others were on the fiber that was in perfect condition. The services that were working on the open fiber presenting the alarm LP_RDI (8 services) and services that were working on the fiber that had no problem presenting the alarm TU_AIS (3 services).
We did the following procedure:
* Check the valid route of the services affected.
* We changed the way in which they were working (force swith). When we did this action with services which presented LP_RDI, the alarm LP_RDI disappeared but the alarm TU_AIS appeared and same way with services which presented TU_AIS, the alarm disappeared but the alarm LP_RDI appeared.
* Modify the trail. it didn?t work.
* We changed the SNCP cross-connections by one bidirectional cross-connection on some of the nodes in the ring to force the switching but the service continued alarmed.
* We did a loop on node C (direction of node D) and service in the virtual network was normal.
We found that problem was on virtual NE, in order to normalize the services afected, we should to commute the working channel over the virtual NE, this action could mean the service affectation of the other services, logically, this action did not be executed.
The real problem was the following:
* The first one is that in nodes A, B and C, the cross-connections are created as Vc12 (low order) and over the node D (virtual NE) the cross-connections are created as Vc4 (high order).
* The second one is that when a fiber cut occur between nodes AB or BC, the virtual NE does not switch, because it does not receive any alarm, therefore, the services which needs that the virtual NE does its switch, are affected.
We have a ring comprising for 4 NE, its protection is PP, one NE is a virtual equipment, when a fiber cut occur, most of the services commute, but some services are affected and its switched is not appropriately.
Before the fail, all services were working only for one path, this is because the virtual NE can only be as active (working) one of its sides.
The fiber cut ocurred
Our recomendation is the following:
* Configure the same type of cross-connections on the virtual NE, you must change the Vc4 cross-connections by Vc12 cross-connections.
* You must change the PP protection by MSP protection.