I am running a average sized wireless network. (9 AP’s & 28 Stations). I am running OSPF across the whole WAN. It has been running fine for about a month now all of a sudden, on random links the OPSF is coming up with
state=Init
and the traffic is not routed through. The only way it can be resolved is if i reboot the remote routerboard then it works again for a few days. I am running v2.8.26 on all the boards. Any help would be appreciated.
We are seing the same issue but it happens in about 10 minutes for us. The OSPF links to Full then about 10 minutes later it drops to INIT.
Does anyone have an answer for this?
happened to me too - seems 2.8.26 is buggy.
I have about 20 ospf MT connected routers in my network and can confirm that with 2.8.26. Previous versions run OK.
BTW do telnet 0.0.0.0 in command promt and benefit from all quagga can offer 
Ports are 2604 - ospf, 2605 - bgpd.
Cheers!
Thanks for the information on how to get into quagga. Never thought to try that and it gives us information that is much needed.
please - everyone that has encountered this problem, please make a supout.rif file at the moment the OSPF is not working, and - with a description of the problem and your network setup, please send it to support
i just mailed a supout.rif file..
i hope this can help..
any news ?
i noticed that this happens between 2.8.26 routers.
I have a 2.8.26 box connected with 5 other MT boxes, running different versions, and ospf hungs only on one interface, this one is connected to a 2.8.26 also..
I am having the same issue with a .26 version running ospf with a cisco router. Had no problems with older version, now it only goes to init on the cisco side and init on routeros side.
Help…
Update, it looks like the routeros is not seeing its own router id in the hellos being send back from the other router.
So its not getting past the init state.
keep sending them supout’s
anyone has 8.2.25 all-packages ?
i have to switch back.. our network is a mess with this bug…
2.8.25 on one side, 2.8.26 on the other,
same sh#t…
i’ll try to downgrade the second one…
2.8.25 on one side, 2.8.25 on the other,
same sh#t…
do i have to downgrade to 2.7.xx ?
We have this problem, but not so often ( once in a month or two months). You can write a sript tat checks route table ,and if ts too short, do a reboot. Ops, we use 2.8.19, and as I see we won’t upgrade till 2.9.
I’m talking about a community wireless network with more than 40 MT routers, everyone has 2-5 wireless interfaces (running ospf)..
and 250+ users…
Since this is a closed network, without internet access, every file transfer occurs between users..
Our main routers has 12-15Mbps traffic in EACH interface…
thats why i feel anxious about this bug…
a new version is out, 2.8.27 but this is the only comment in change log is:
What’s new in 2.8.27:
*) added optional wireless package for RouterBOARD R52 (Atheros 5414/13) support;
what about ospf ?
Very interesting, but…
What is the password?
Quagga does not appear to have a default, nothing obvious worked, and the system admin password did not work.
Mostly, I’m just interested in looking around, not changing anything. But, the enable password would be cool too.
Thanks,
–Eric
Well, quagga is based on zebra
Anyhow, misconfigure and the thing blows up!
Gotcha, Thanks!
Something to be very carefull with, but quite interesting to look at.