Hi, everyone! Today I’m going to introduce to you how to deal with high and low CPU cases.
Issue Description
NE40E CPU is high sometimes.
Alarm Information
Handling ProcessFrom the current information, the CPU high because of that there are too many SNMP request to NE40E, you added too many nodes on NMS. Please delete some monitoring task on NMS. If can’t, please filter those IP address on NE40E to reduce the CPU usage. The details are as follows.
1. When CPU usage is high, the CMF and IFM process is high. Those two process high means that there are too many NMS tasks collecting interfaces information and make CPU high.
2. From the process, there are many MTif script operation, it means there are too many NMS tasks collecting interfaces information.
[NE40E-diagnose]disp cmf luascript history p 3 *************************************BEGIN************************************* The history script from ecm: <?xml:namespace prefix = "o" />TransNo SessionID Optype MsgType pSptCb RunNum RunRetCode SptRetCode TotalMem(K) CurSptMem(K) LoadTime(ms) MicroRunTime(ms) SptFileName StartRunTime FinishRunTime CancelTime TimeOut LuaLastErrInfo 0x14e62403 0x0000028b 0x00000009 0x00000005 0x5c44228c 4 0x00800910 0x00000000 98391 48 0 20 MTIf.lua 2017-10-30 22:43:47.752+05:30 2017-10-30 22:43:47.768+05:30 0000-00-00 00:00:00:000 -0001-11-29 16:00:000 0x14e62404 0x0000028b 0x00000009 0x00000005 0x5c44228c 4 0x00800910 0x00000000 98441 50 0 10 MTIf.lua 2017-10-30 22:43:47.769+05:30 2017-10-30 22:43:47.786+05:30 0000-00-00 00:00:00:000 -0001-11-29 16:00:000 0x14e62405 0x0000028b 0x00000009 0x00000005 0x5c44228c 4 0x00800910 0x00000000 98490 47 0 10 MTIf.lua 2017-10-30 22:43:47.786+05:30 2017-10-30 22:43:47.803+05:30 0000-00-00 00:00:00:000 -0001-11-29 16:00:000 0x14e62406 0x0000028b 0x00000009 0x00000005 0x5c44228c 4 0x00800910 0x00000000 98541 51 0 10 MTIf.lua 2017-10-30 22:43:47.806+05:30 2017-10-30 22:43:47.824+05:30 0000-00-00 00:00:00:000 -0001-11-29 16:00:000 ...... |
3. From the logs, some source IP sent many pakcets to NE40E, expecially 20.x.x.5和20.x.x.10.
| [NE40E_-diagnose]disp snm dia ac ------------------------------------------------------------------------------------------------------------------------------------ Last Accessed Users ------------------------------------------------------------------------------------------------------------------------------------ NMS-IP/Port/VPN Access-Time Pkt-Len Count ------------------------------------------------------------------------------------------------------------------------------------ x.x.x.x7/42372/0 2017-10-30, 22:51:01:041+05:30 825 16982 20.x.x.5/59959/0 2017-10-30, 22:51:00:115+05:30 102 46427427 y.y.y.y/49630/0 2017-10-30, 22:51:00:096+05:30 78 42629 20.x.x.10/62853/0 2017-10-30, 22:50:59:338+05:30 122 1657930 y.y.y.y/44325/0 2017-10-30, 22:50:47:691+05:30 43 2492 …… | [NE40E_-diagnose]disp snm dia ac ------------------------------------------------------------------------------------------------------------------------------------ Last Accessed Users ------------------------------------------------------------------------------------------------------------------------------------ NMS-IP/Port/VPN Access-Time Pkt-Len Count ------------------------------------------------------------------------------------------------------------------------------------ x.x.x.x/42372/0 2017-10-30, 22:51:08:741+05:30 825 17054 20.x.x.5/60058/0 2017-10-30, 22:51:08:116+05:30 102 46427468 …… 20.x.x.10/62853/0 2017-10-30, 22:51:01:844+05:30 118 1657938 …… |
Root CauseThe CPU high because of that there are too many SNMP request to NE40E.
SolutionDelete some tasks on NMS or shorten the intercal of collecting data.
Any further questions, let us know!
