Showing results for 
Search instead for 
Did you mean: 

Cisco ITP SCTP M2PA link goes down


I am having an issue with a M2PA linkset going down like 10 times a day for a period of 20s. I believe this can be fixed by tuning the timeouts but I am unable to find the good commands.

This is the log:

Jan 24 14:14:43.814 GMT: %CS7M2PA-3-DESTADDRINA: Destination IP Address inactive for linkset:slc Linkset0:0

Jan 24 14:14:43.822 GMT: %CS7MTP3-3-LINKNOTUP: Link 0 in linkset Linkset0 failed. Reason=communication lost

Jan 24 14:14:43.822 GMT: %CS7MTP3-5-LINKUPDOWN: Link 0 in linkset Linkset0 is unavailable

Jan 24 14:14:43.834 GMT: SCTP: Invalid association state for sctp_send

Jan 24 14:14:44.623 GMT: %CS7MTP3-3-LINKSETSTATE: Linkset Linkset0 is unavailable

Jan 24 14:14:44.972 GMT:  DEAD   snmpAssocP:0x83B088EC assocP:0x8397EC84 Id:0x002C0151 dc:  17 avl:  18

Jan 24 14:14:45.044 GMT: SCTP: Out-of-the-blue datagram rcvd from addr, port 4444, discarding...

Jan 24 14:14:45.048 GMT: SCTP: Out-of-the-blue datagram rcvd from addr, port 4444, discarding...

Jan 24 14:14:59.973 GMT:  ENTRY DELETE_OLD: Force:False delta:52910643 purgeTime:86400 MaxHeld:100

Jan 24 14:14:59.973 GMT:                  DeadCount:18 AvlCount:18

Jan 24 14:14:59.973 GMT:      entriesProcessed:18 deleteByCount:0 deleteByTime:7

Jan 24 14:14:59.973 GMT:  CREATE snmpAssocP:0x83B01E5C assocP:0x8397EC84 Id:0x002C0152 dc:  11 avl:  12

Jan 24 14:15:01.888 GMT: %CS7MTP3-5-ADJRBEGN: Adjacent SP restart beginning for linkset Linkset0

Jan 24 14:15:02.169 GMT: %CS7MTP3-5-LINKUPDOWN: Link 0 in linkset Linkset0 is available

Jan 24 14:15:02.413 GMT: %CS7MTP3-5-ADJRENDS: Adjacent SP restart ending for linkset Linkset0

Jan 24 14:15:02.413 GMT: %CS7MTP3-3-LINKSETSTATE: Linkset Linkset0 is available

show cs7 m2pa sctp parameters Linkset0

** SCTP Association Parameters AssocID:0x002C0158

AssocID: 0x002C0158 Instance ID: 0 Offload: No

Assoc state: ESTABLISHED  Context: 2208044432 Uptime: 05:20:14.928

Local port: 4444

Local addresses:

Remote port: 4444

Primary dest addr:

Effective primary dest addr:

Destination addresses:   State:  ACTIVE

Heartbeats:  Enabled   Timeout: 300 ms

RTO/RTT/SRTT: 330/257/270 ms   TOS: 0  MTU: 1500

cwnd: 3044  ssthresh: 3000  outstand: 908

Retrans cwnd rate: 50  Retrans cwnd mode: RFC

FastRetrans cwnd rate: 50  Idle dest cwnd rate: 50

Num retrans: 22  Max retrans: 4  Num times failed: 0 retrans: 0

Local vertag: 7102ABE7  Remote vertag: 71D1A261

Num inbound streams: 2  outbound streams: 2

Max assoc retrans: 10  Max init retrans: 8

CumSack timeout: 200 ms  Bundle timeout: 5 ms enabled

Min RTO: 250 ms  Max RTO: 500 ms

LocalRwnd: 64000  Low: 58285   RemoteRwnd: 63092  Low: 61824

Congest levels: 4  current level: 0  high mark: 38  chkSum: crc32

In bold, the timeout of the hearbeats seems incredibly low compared to my RTO. But I dont find the command to increase it.




The heartbeat timer in SCTP is used to check availabity of the dual homed / peer IP address thus should not affect stability of your SCTPassociations

your M2PA links are between which two elements? you should compare the timers betwwen the two elements! and provide both the timers and linkset/link configurations