[c-nsp] 6500, SUP720, 12.2(33)SXI3, BGP path-mtu flap
Tim Durack
tdurack at gmail.com
Thu Feb 25 13:01:40 EST 2010
6500, SUP720, 12.2(33)SXI3.
Couple of iBGP neighbors keep flapping on one box. Looks like a
path-mtu issue, yet the discovered path looks correct:
sh bgp ipv4 unicast summary | i Neighbor|10.116.0.1
Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ
Up/Down State/PfxRcd
10.116.0.1 4 65116 165937 690347 268786 0 61
00:00:50 0
path-mtu discovers 9116:
Address tracking is enabled, the RIB does have a route to 10.116.0.1
Connections established 3937; dropped 3936
Last reset 00:02:22, due to BGP Notification sent, hold time expired
External BGP neighbor may be up to 255 hops away.
Transport(tcp) path-mtu-discovery is enabled
Connection state is ESTAB, I/O status: 1, unread input bytes: 0
Mininum incoming TTL 0, Outgoing TTL 255
Local host: 10.1.0.1, Local port: 46013
Foreign host: 10.116.0.1, Foreign port: 179
Enqueued packets for retransmit: 2, input: 0 mis-ordered: 0 (0 bytes)
Event Timers (current time is 0x146C10A6C):
Timer Starts Wakeups Next
Retrans 9 5 0x146C121A4
TimeWait 0 0 0x0
AckHold 1 0 0x0
SendWnd 0 0 0x0
KeepAlive 0 0 0x0
GiveUp 0 0 0x0
PmtuAger 1 0 0x0
DeadWait 0 0 0x0
iss: 3884779463 snduna: 3884779560 sndnxt: 3884788712 sndwnd: 16288
irs: 1703625207 rcvnxt: 1703625304 rcvwnd: 16288 delrcvwnd: 96
SRTT: 99 ms, RTTO: 1539 ms, RTV: 1440 ms, KRTT: 49248 ms
minRTT: 8 ms, maxRTT: 300 ms, ACK hold: 200 ms
Flags: higher precedence, nagle, path mtu capable, path mtu discovery
Datagrams (max data segment is 9116 bytes):
Rcvd: 4 (out of order: 0), with data: 2, total data bytes: 96
Sent: 5 (retransmit: 5), with data: 3, total data bytes: 9248
Pinging between loopbacks with df-bit and mtu 9156 is successful.
Disabling path-mtu on the neighbor session "fixes" the problem. The
problem only started occurring after some outages caused the sessions
to flap.
Anyone seen odd path-mtu related issues?
--
Tim:>
Sent from New York, NY, United States
More information about the cisco-nsp
mailing list