[rbak-nsp] Strange traffic breakdowns RB

Dariusz Siedlecki siedar at pronet.lublin.pl
Tue Mar 11 18:17:43 EDT 2014


Hello Redback Users,

We have strange hangups - BGP sessions going down,  Radius loosing 
connection.

I have strange breakdowns on traffic - some times once a day, sometimes 
3 times in rough hours.

SE600, XCRP 4,   10ge-4-port, SEOS 6.2.1.9

It appears from about last two weeks  and appears very randomized 
(sometimes every day, sometimes 2 times a week)

Always I have problems with lost connection to radius and usually 
bgp/isis stops responding at that moment.

Mar 10 20:30:12: %SYSLOG-6-INFO: aaad: Radius Auth srv 192.168.89.2/1812 
(1812) no response in 31 sec. (user 00:02:44:30:ec:9f).Marked dead
Mar 10 20:30:54: %SYSLOG-6-INFO: aaad: Radius Acct srv 192.168.89.2/1813 
(1812) no response in 69 sec. (user 00:23:54:6b:2a:39).Marked dead

Some times  I have problem with BGP sessions also (some times just one 
or two sessions)

Mar 10 20:37:47: [0002]: %BGP-6-INFO: 10.10.10.50 DOWN - Notification sent
Mar 10 20:37:47: [0002]: %BGP-6-INFO: 10.10.10.50 send NOTIFICATION: 4/0 
(hold time expired) with 0 byte data. mxReadMs=30048
Mar 10 20:37:55: %SYSLOG-6-INFO: aaad: Radius Authentication srv 
192.168.89.8/1812 (1812) received response (user Redback).Marked Alive
Mar 10 20:38:24: [0002]: %BGP-6-INFO: 10.10.10.50 rcv NOTIFICATION: 6/5 
(cease: connection rejected) with 0 byte data. mxWritMs=30011
Mar 10 20:38:58: [0002]: %BGP-6-INFO: 10.10.10.50 UP

Noting more in logs.

We are using BGP, Clips interfaces, and ISIS between JuniperMX/Huawei9300.

Show crashdumps are empty.

We have up to 6Gb/s in peak and about 500Kpps - it happens between 7pm 
till 9pm.

Maybe someone have similar problem?

I think that it is related with some type of flood or strange type of 
packet coming to the RB.

Any ideas?

Best regards

Dariusz Siedlecki






More information about the redback-nsp mailing list