A carrier delivers NQA configurations to devices in the city to test their delays. After NQA is configured on the NE80E or NE40E to send ICMP packets to test the delays on the SR, BAS, and SW devices, the delay is long. In actual ping tests, the delay is not long.
The NQA configuration on the NE80E is as follows:
nqa test-instance ipnet SDYATMCRT06
destination-address ipv4 xx.xx.xx.xx
interval seconds 1
The delays on tested devices are long.
1. The link quality is poor, and the long delay is caused by physical factors.
2. The NE80E or NE40E is incorrectly configured.
3. NQA instances on a tested device affect each other and cannot accurately reflect the delay.
1. Ping the system software of the tested device to test link quality. The delay is not long, indicating that link quality is normal.
2. Check the NQA configuration according to the command reference. The NQA configuration is correct.
3. Delete an existing NQA instance and configure it again. The delay is greatly decreased, and the problem is solved.
A large number of NQA instances are simultaneously configured. All NQA instances send packets at the same time. The NE80E or NE40E cannot process Echo Reply messages in a timely manner, leading to long delays.
Run the display nqa results command to check the instances with a long delay or use the NMS to import devices with long delays and find instances to which the devices belong. Enter and stop the instances. Wait for 1 minute, and run the start now command. The problem is solved.
On the live network, after scripts are edited, the scripts are delivered on the NE80E simultaneously so that the startup time of NQA test instances are the same. The NQA module sends ICMP packets simultaneously. Consequently, the NE80E cannot process ICMP packets in a timely manner. There is a possibility that the delay is large. Configure NQA test instances at different times so that ICMP packets are not sent simultaneously.