Chinese
English
Description of the problem
The customer configures EPL service on EMS4 board, and VCTRUNK binds 5 VC12 services to two upper fibers respectively. After enabling LCAS function, when fiber breakage occurs, the data service is interrupted for about 150ms.
Customers ask us what is the protection reversal time of LCAS in this scenario, is there any standard?
Processing
1. Standard time calculation
The multi-frame of VC12 is realized by the b2 bit of K4 bytes, 32 VC12 multi-frames form a virtual cascade multi-frame, and each VC12 multi-frame consists of 4 VC12s. Therefore, the period of the virtual cascade compound frame composed of VC12s is 32*4*0.125ms = 16ms
According to the standard description, the recovery time of the MFI can be 3~10 virtual cascade reframe cycles.
The MFI recovery time of VC12 is that:
3 MFI recovery re-framing cycles = 3 * 16ms = 48ms
10 MFI recovery reframe cycles = 10 * 16ms = 160ms
2. LCAS protection reversal time = fault detection time + single board service recovery time + LCAS troubleshooting time.
1) The fault detection time is related to the actual path length of the existing network.
2) The single board service recovery time is between 48ms~160ms according to the standard definition when VC12 is configured.
3) LCAS fault processing time,The LCAS task processing period of EMS4 is 200ms, which means the maximum processing time of LCAS task is 200ms.
4) Combined with the above calculation, the LCAS reversal time is around 200ms.
Root cause
In ITU-T, the LCAS standard does not define the inversion time for LCAS protection, however, the LCAS protocol itself is based on the virtual cascade technology, and the virtual cascade technology standard has strict regulations on the service recovery operation in order to ensure the correct service recovery.
The low-order MFI recovery time is calculated as follows:
G.783 standard description:
8.2.5.2 VC-11 VC-12 VC-2 virtual concatenation multiframe alignment
Sink direction The extended overhead multiframe is recovered according to 8.2.3.1. The virtual concatenation frame number is recovered from bits 1-5 of the K4[2] multiframe sequence. The process will be in the out of multiframe (OOM) state when either the extended overhead multiframe alignment process is in the OOM or the extended overhead multiframe alignment process is in the OOM. in the OOM
The process will be in the out of multiframe (OOM) state when either the extended overhead multiframe alignment process is in the OOM state or when an error is encountered in the received and expected frame number from bits 1-5 of the K4[2] sequence. extended overhead multiframe process is in the IM state and two consecutive error-free frame numbers are recovered. The sequence number is recovered from bits 6-11 of the K4[2] sequence. The sequence number is recovered from bits 6-11 of the K4[2] sequence. A new sequence number is accepted if the received sequence has the same value in n consecutive extended overhead A new sequence number is accepted if the received sequence has the same value in n consecutive extended overhead multiframes, with 3 ≤ ?n ≤ ?10. The accepted sequence number is compared with the expected sequence number for detection of the dSQM defect.
Recommendations and Summary
The LCAS function is lossless in the scenario of adding and deleting bindings (provided that the actual traffic does not exceed the bandwidth size after deleting bindings or adding bindings). And it can't do lossless protection for the scenario of interrupting fiber.
