High CPU usage peak on NE40E router

Issue Description

The NE40E router is showing high CPU usage peaks.

The NE40 router is showing high CPU usage peaks

Handling Process

1. performed the command “display cpu-usage history 24” around 9:30 am 9th Feb:

9th Feb

So, the peak happened at around 23:30 pm on 8th Feb. we have confirmed, in this picture, from the left to the right, the time is from now to the lasted 24H. so the peak happened at 10 hours early than we run the command.

2. The peak is not 100%, it means in the range from 95% to 100%. Check the logs, the peak is 99%. So will not update the Max CPU Usage.

Max CPU Usage

3. Check the history alarms on the device, there are no high CPU usage alarms, so, it means, the CPU usage was high only for a short moment. It’s not keeping high. So, please don’t worry about it. Has no impact.

 

4. Check the logs, we find the high CPU usage was caused by BGP flapping, and the device refresh the routes:

Line 115791: Feb  8 2022 23:36:09-03:00 NE40E %BGP/6/NOTIFY_RECV(l): CID=0x801304a8-Slot=4; The router received a NOTIFICATION from the peer. (Peer=10.x.x.85, SourceInterface=LoopBack0, ErrorCode=4, SubErrorCode=0, NotifyInfo=Hold Timer Expired,VpnInstance=_public_, ErrorData=NULL)

Line 115800: Feb  8 2022 23:36:11-03:00 NE40E %BGP/6/NOTIFY_RECV(l): CID=0x801304d1; The router received a NOTIFICATION from the peer. (Peer=10.y.y.5, SourceInterface=LoopBack0, ErrorCode=4, SubErrorCode=0, NotifyInfo=Hold Timer Expired,VpnInstance=_public_, ErrorData=NULL)

Line 115826: Feb  8 2022 23:36:42-03:00 NE40E %BGP/6/BNM_PEER_SWITCHIN_SENDING_LIST(l): CID=0x801304d1; The BGP peer entered the batch advertisement queue. (PeerAddress=10.x.x.85, VPNInstance=_public_, AddressFamily=FLOWv4, UptPeerGrpID=5)

Line 115833: Feb  8 2022 23:36:45-03:00 NE40E %BGP/6/BNM_PEER_SWITCHIN_SENDING_LIST(l): CID=0x801304d1; The BGP peer entered the batch advertisement queue. (PeerAddress=10.y.y.5, VPNInstance=_public_, AddressFamily=IPv4-unicast, UptPeerGrpID=3)

Root Cause

BGP down/up and device refresh routes, cause high CPU usage in a short moment.

Solution

Its normal behavior, BGP down/up and device refresh routes, cause the high CPU usage in a short moment. There were no high CPU usage alarms, so, it means, the CPU usage was high only for a short moment. It’s not keeping high. So, please don’t worry about it. Has no impact.

Suggestions

By the command: display CPU-usage history 24, we can check the CPU usage for the last 24  hours. Sometimes, the CPU usage will be high in a short moment. If there aren’t CPU usage alarms, it has no impact.