-----Original Message----- From: Public [mailto:public-bounces@talk.mikrotik.com.au] On Behalf Of Paul Julian Sent: Friday, 30 January 2015 7:59 AM To: 'MikroTik Australia Public List' Subject: Re: [MT-AU Public] Strange routerboard behavoir
Hi Mike, it seems to be the AP that is playing up, although in saying that I am not 100% sure because it just starts working, I've never really been able to do anything that actually gets it to work except a reboot of the AP, so realistically it could be the AP or CPE but I think it's the AP as I can still get into
wireless.
Some devices are 5.25 still, others 6.11 or higher, the AP and CPE always match versions also.
The wireless from the CPE to the backhaul AP is using station mode, it's a routed network and the wireless backhaul network is solid, I haven't seen one drop in that since the last reboot of the whole network 2 months ago.
We had another one do it yesterday, just dropped off then came back 3 times, first time disappeared for 3h25m, second time 6h42m, 3rd time 30m, so now I can get back into it I can see it's been up for 59 days, the log on the CPE shows nothing happening at the time the AP dropped off, the log of the AP also shows nothing in the logs, as if we never lost contact with it and it never had a problem.
Strange....
Regards Paul
-----Original Message----- From: Public [mailto:public-bounces@talk.mikrotik.com.au] On Behalf Of Mike Everest Sent: Thursday, 29 January 2015 3:51 PM To: 'MikroTik Australia Public List' Subject: Re: [MT-AU Public] Strange routerboard behavoir
Hi Paul,
Is this what we are talking about?: 5GHz-base-station <---> CPE <---> 2GHz- AP
And the 'CPE' is the one that behaves badly in the manner described?
What software version is on each one, and how is it configured? Is the CPE mode station or station bridge or something else? And what interface is
HI Mike, yes routing mode, none of the addresses respond when the problem happens, even UDP doesn't work, so no SNMP either. Pings even from the directly attached backhaul router don't get a response but the port state is running and that looks OK. Mac telnet from the backhaul CPE doesn't work either, just says disconnected, it does try and asks you for a username to login but once you enter that it just comes up straight away with disconnected. I haven't actually tried a torch, I agree it's a great tool and I don't know why I haven't tried it....or maybe I have and just can't remember, but will try it again :-) I will check ARP tables as well, I don't recall checking that though. Thanks for the suggestions Mike ! Regards Paul -----Original Message----- From: Public [mailto:public-bounces@talk.mikrotik.com.au] On Behalf Of Mike Everest Sent: Friday, 30 January 2015 11:06 AM To: 'MikroTik Australia Public List' Subject: Re: [MT-AU Public] Strange routerboard behavoir I see, So if it is a routing node, I suppose it has a different address on each interface - do you get the same result on both of those addresses? You still get ping reply from that address when the problem happens, but you can't connect with winbox or webfig, telnet or mac-telnet? Sounds like you have some kind of SNMP monitoring going on, so do pings from the management system and from your local admin PC behave the same way? Have you tried looking at torch tool to check for traffic on each interim router? Have you checked ARP tables on adjacent routers and maybe bridge host tables? Maybe there are one or more interfaces with proxy-arp enabled that could cause this sort of weirdness? Or bridging table loop that confuses a forwarding process? I always like torch tool in situations like this - if there is any kind of incorectly forwarded or natted packets, torch will offer up useful clues every time! :-D Cheers! Mike. the CPE via the
IP attached to?
Cheers!
Mike.
-----Original Message----- From: Public [mailto:public-bounces@talk.mikrotik.com.au] On Behalf Of Paul Julian Sent: Thursday, 29 January 2015 12:59 PM To: public@talk.mikrotik.com.au Subject: [MT-AU Public] Strange routerboard behavoir
Hi guys, we have a large hotspot site which is doing some funny things and I was just wondering if anybody may have seen similar behaviour before.
We have a 5Ghz backhaul network which runs flawlessly, running with a mix of SXT's and Omnitiks from a central AP out to a number of CPE's. Off the back of each CPE we run a 411 with a 2.4Ghz radio in it and an omni antenna, plugged into ether1 on each device. What happens sometimes is that we lose sight of the AP connected to a CPE, we can't ping it, however the AP shows in the neighbours list on the CPE and the Ethernet port shows as running, for all intents and purposes everything looks fine, however when you try to telnet from the CPE to the AP you can't connect, when you try and mac-telnet from the CPE to the AP it disconnects and doesn't work.
After a varying period of time the connection just starts working again, the uptime of the AP is normal, i.e. it hasn't lost power or rebooted and everything just works OK....until the next time which could be hours or days, no pattern seems to exist.
Has anybody encountered such issues before at all ? I have so far replaced the Ethernet joiner between the radios and also replaced the power pack and POE injectors which feed the radios.
Any suggestions would be appreciated.
Thanks Paul _______________________________________________ Public mailing list Public@talk.mikrotik.com.au http://talk.mikrotik.com.au/mailman/listinfo/public_talk.mikrotik.com. au
_______________________________________________ Public mailing list Public@talk.mikrotik.com.au http://talk.mikrotik.com.au/mailman/listinfo/public_talk.mikrotik.com. au
_______________________________________________ Public mailing list Public@talk.mikrotik.com.au http://talk.mikrotik.com.au/mailman/listinfo/public_talk.mikrotik.com. au
_______________________________________________ Public mailing list Public@talk.mikrotik.com.au http://talk.mikrotik.com.au/mailman/listinfo/public_talk.mikrotik.com.au