Hello i have implemented OSPF on my network, the type is PTMP with 3 routers.
The problem is that i have a lot of little downs of neighbors.
The links between routers is by wireless all of them running week by week without downs.
In the attached screenshot show log of main router when the problem dissconect neighbors.
Thank to all.
Hey. Check your router-id’s on all routers. Are they unique?
Yes it´s unique
Run this and paste it here.
/export hide-sensitive
This will give us a much better idea of what exactly you have built.
1 is the main router and 2 is another secondary router.
Ok, are you getting these links downs from all the remote routers?
In you log you posted I see router ID 10.254.254.2 but you have posted the export for 10.254.254.4. Is router 2 that you posted the config for, a remote router?
Is these a good reason you have forced area 1 on router 1 but not forced it on router 2?
Its a little annoying but do the following commands on router 1 and 2 and post the results
routing ospf neigh print
routing ospf interface print
routing ospf network print
routing ospf lsa print
routing ospf area print
Also ping from one router to the other let it run for say 1000 pings. and post those results.
Run something like this
ping 192.168.11.1 interval=100ms count=1000
and post the result something like this.
sent=1000 received=1000 packet-loss=0% min-rtt=19ms avg-rtt=29ms max-rtt=44ms
Perhaps a drawing showing the layout of your network might help a lot here.
I attach all the information, if you need more information or anything let me know.
Thank you very much for your help.
socada.rar (41.8 KB)
Ok so I’ve had a good look at the configs, the ospf prints and can’t see anything dangerously wrong.
A few more questions if you could.
What is the difference between a solid and dotted black lines on your topology?
I see the adjacencies (neighbor.txt) have been up for some hours, what sort of interval to they typically last?
Routers 2 and 3 look like they have formed an adjacency as well as with router 1. Did you intend for that to happen?
In the ospf interface setup is there a particular reason you went for a broadcast interface that’s passive and then added NBMA neighbors?
On router 2 you used the bridge interface for OSPF and on router 3 you used ether6 for OSPF, is that intentional?
Run “routing ospf neigh print” again on router 2. See the adjacency time back to 10.254.254.1 as shown below.
Check the logs of your other devices from Router 1 to the switch with the blue circle below it. See if something happened on those devices at the time of the adjacency change.
0 instance=default router-id=10.254.254.1 address=172.22.2.1 interface=bridge1
priority=1 dr-address=0.0.0.0 backup-dr-address=0.0.0.0 state=“Full”
state-changes=26 ls-retransmits=0 ls-requests=0 db-summaries=0
adjacency=9h37m
Also on Router 1 open up winbox and go to tools > profile and then hit start. Watch it for a few minutes and attach a picture of it’s normal behavior.
Ok I answer in the order you ask.
0 instance=default router-id=10.254.254.1 address=172.22.2.1 interface=bridge1 priority=1 dr-address=0.0.0.0 backup-dr-address=0.0.0.0 state=“Full”
state-changes=26 ls-retransmits=0 ls-requests=0 db-summaries=0 adjacency=12h39m14s
And I check the switch.
Log when OSPF goes down on router 2
18:50:00 route,ospf,info OSPFv2 neighbor 10.254.254.1: state change from Full to Down
18:50:01 route,ospf,info Ignoring Link State Acknowledgment packet: wrong peer state
18:50:01 route,ospf,info state=ExStart
19:15:00 route,ospf,info OSPFv2 neighbor 10.254.254.1: state change from Full to Down
19:15:01 route,ospf,info Discarding packet: no neighbor with this source address
19:15:01 route,ospf,info RouterId=10.254.254.1
19:15:01 route,ospf,info source=172.22.2.1
profile.txt (4.84 KB)
logOspf.txt (48.1 KB)
Sorry for the slow reply, i’ve been swamped lately.
I assume the devices I’ve circled are yours?
When OSPF goes down, can you check the logs on these to see if anything happened at the same time?
Next up is to probably turn logging on, it will generate significant log entries. You probably need to capture one ospf event and 10 minutes either side.
Then attach in your reply.
/system logging
add action=disk topics=ospf
I would suggest running a long term packet capture only for IP Protcol 89 on each side of the link so that we can look at the packets as it’s failing. You may need to let it run for a while so be sure to give it enough memory and only allow OSPF packets which are small.
That’s how I normally troubleshoot complex OSPF issues for the clients I help with MikroTik.
Thanks to all for your help, this weekend I was working on this and change the type to NBMA, I try with this type time ago but fail too. But now working fine by 3 days.
I think that time ago when try NBMA fail by a misconfiguration. Now I configure good all and priorities that I think was the problem the last time I used.
Really NBMA is the type I wanted to use first, all my links are wireless and so I delete broadcast.
Really thanks to all.