Five OSN 3500 ECC communication interrupted intermittently. NE's intermittently log off and log in without graying out. Sending new configurations also fails to active the cross connections in the NE's which have the communication issue.
When create service, operation fails. Error: User has not logged in. See screenshot in attachement
1. Unlike other ECC faults this time the 3 * OSN 3500 NE's do not grey out on topology.
2. Select NE --> right click and select Ping --> is successful.
----10.25.2.55 PING Statistics----
4 packets transmitted, 4 packets received, 0% packet loss
round-trip (ms) min/avg/max/stddev = 5.95/6.39/6.60/0.30
3. Check ECC route: Use :cm-get-eccroute; Confimred the offline nodes can be seen; also confirmed from U2000 --> NE Explorer -- > Communication --> ECC link management. (Note: very many NE's seen because require optimization)
DST-ID DXC-ID DISTANCE LEVEL MODE SCC-NO PEER-SCCNO
0x000901f6 0x0009015c 6 4 auto 2 0
0x000901f7 0x0009015c 5 4 auto 2 0
0x000900e6 0x0009015c 5 4 auto 2 0
0x000900e7 0x0009015c 4 4 auto 2 0
0x000900e9 0x0009015c 4 4 auto 2 0
Total records :146.
The number of ECC routes is very many and Huawei recommended number is a maximum of 60 NE to one GNE. The ECC requires optimization. However this is not the root cause since the large ECC subnet has been existing for a while and ECC subnet size impacts all NE’s in the form of ECC broadcast storms. For this case only three OSN affected also only intermittently.
6. Manually trace the ECC route starting from the GNE from NE Explorer --> Communication --> ECC link management and check all the NE which distance is 0. From here trace further and check for next hop until arrive at disconnecting NE and mark the line card being used. See screenshot in attachment.
7. Check the line card (board) currently used for ECC communication. Browse alarms on this board find that there are B1_SD and B2_EC alarms occur intermittently affecting the communication link. See screenshot in attachment.
8. Disable the DCC (D1-D3) bytes on this line card and the ECC route auto refresh and stabilize. Another method is to create a manual route however please note the manual ECC route has a higher priority than the auto route and can cause other ECC problems if configured wrongly. See screenshot in attachment.
The auto created ECC route uses a link which has very many B1_SD and B2_EC intermittent alarms. As a result every time the alarms occur, (errors occur) the ECC link cannot be established. See handling process below for more information.
Ensure errors are eliminated or kept to a minimum in the network.