Page 1 of 1
Clients loosing connections in iregular intervals on 2.8.17
Posted: Fri Oct 15, 2004 9:40 am
by spavkov
I have strange problem...
Im running hotspot with universal client and enabled-address method and have 13 (!!!) users....
i also use pheenet 1watt amplifier and 15db omni antenna....
some of my clients that have good signal strenght are loosing connection in intervals of 5-10 minutes...
Client lose his connection and in 4-5 seconds he gets connected again.
sometimes i see lines in log like this:
17:03:40 client XXXX desconnected.
17:03:43 client XXXX connected.
17:03:50 client XXXX desconnected.
17:03:56 client XXXX connected.
17:04:10 client XXXX desconnected.
17:04:11 client XXXX connected.
What could be causing this problem... can it be universal client?
because i used before hotspot without universal client and it worked ok...
But i need universal client so that my users that have AP instead of wireless card can connect on hotspot...
any ideas...???
Posted: Fri Oct 15, 2004 12:53 pm
by tully
Please send a supout file to
support@mikrotik.com
John
Thank you
Posted: Fri Oct 15, 2004 1:05 pm
by spavkov
Thank you for prompt repy.
I will send support file with detailed description of problem at 15h when i get home...
Greets
Slobo
Exactly the same problem, instead im using a P-P link
Posted: Wed Oct 27, 2004 7:48 am
by ripunjay
hi, im facing the same problem, only diff, im using a point to point link, prism card 200mW
Earlier were using MT 2.6 earlier, with exactly the same setup, without any problems for almost a year. The machine is an AMD. other side we have a senao wireless bridge. distance is +/- 10Km
support.rif sent
Posted: Wed Oct 27, 2004 3:03 pm
by kchris
we have almost the same problem!!! ALL of our wireless clients reconnect after a specific time (a few seconds). We are using 2.8.18.
Posted: Wed Oct 27, 2004 3:21 pm
by tully
If you check the logs, you may see. If you send a supout file, we can see very low level logs and tell you why. It may be that there is water in the cable and that badly affects the physical (RF) part of the radio -- and it can go on and off.
John
Posted: Wed Oct 27, 2004 4:09 pm
by kchris
If you check the logs, you may see. If you send a supout file, we can see very low level logs and tell you why. It may be that there is water in the cable and that badly affects the physical (RF) part of the radio -- and it can go on and off.
John
thank for your reply, my collegaue sent the supout. I don't think that water caused this problem because we switched to Mikrotik AP 2 days ago (from Buffalo WLAs) and before that this problem didn't show up (and the equipment is indoors).
Posted: Wed Oct 27, 2004 10:41 pm
by spavkov
It look that this disconnecting problem happens with clients that have signal of around -60 dbm...
that is what i can figure out...
Posted: Thu Oct 28, 2004 11:28 am
by ripunjay
On upgrading to 2.8.18: the logs show :
Prism firmware loading: skipping, too old primary
Interface Info:
/interface wireless print
Flags: X - disabled, R - running
0 R name="wlan1" mtu=1500 mac-address=00:02:6F:03:16:E1 arp=enabled disable-running-check=no interface-type=Prism
prism-cardtype=200mW radio-name="00026F0316E1" mode=station ssid="XXX" frequency=2422 band=2.4ghz-b scan-list=default-ism
rate-set=configured supported-rates-b=5.5Mbps supported-rates-a/g="" basic-rates-b=5.5Mbps basic-rates-a/g=""
max-station-count=2007 tx-power=default periodic-calibration=default fast-frames=no dfs-mode=none antenna-mode=ant-a
wds-mode=disabled wds-default-bridge=none wds-ignore-ssid=no update-stats-interval=disabled default-authentication=no
default-forwarding=no hide-ssid=yes 802.1x-mode=none
/interface wireless info print
0 interface-type=Prism firmware="1.0.5/1.2.0, compid 0x800c" tx-power-control=yes ack-timeout-control=no alignment-mode=yes
virtual-aps=no noise-floor-control=no scan-support=yes burst-support=no nstreme-support=no supported-bands=2ghz-b
2ghz-b-channels=2412,2417,2422,2427,2432,2437,2442,2447,2452,2457,2462,2467,2472
The RF Link works with a Senao bridge on both the sides without a problem, so Cable problems can be ruled out (hopefully!!!) , but with a senao bridge on remote end and MT 2.8.18 on local side the disconnection problem is continious, and all this happned after the upgrade from 2.6.x to 2.8.x
Posted: Thu Oct 28, 2004 11:33 pm
by sabo
Please upgrade your Prism firmware inside Senao Bridges to version 1.1.1/1.7.4
How to upgrade prism firmware inside senao bridges ???
Posted: Sat Oct 30, 2004 10:57 am
by ripunjay
Can you tell me how to upgrade the firmware inside the senao bridges...
and the error from MT 2.8.18 on bootup is...
Prism firmware loading: skipping, too old primary
where as the prism firmware inside of MT shows up as the following ...
interface-type=Prism firmware="1.0.5/1.2.0, compid 0x800c"
could the problem be with MT side prism card .... ???
thanks
Posted: Sun Oct 31, 2004 1:03 am
by tully
You can write
support@mikrotik.com and they will assist you.
John
Posted: Sun Oct 31, 2004 8:26 pm
by uldis.a
The same problem - clients disconnecting and reconnecting after some seconds some times a day. Atheros cards and 2.8.17. Signals about -60
Posted: Sun Oct 31, 2004 8:30 pm
by uldis.a
Can it be due to some hardware conflicts, or possible attacks? I noticed high CPU usage close to disconnect moments. Hardware is original Intel board with 2 GHz Celeron and 128 Mb RAM. Using 3 Atheros cards on RouterBoard 4 miniPCi-to-PCI adaptor. What else can it be? Any solutions?
Posted: Sat Nov 06, 2004 10:35 am
by savage
I'm starting to pick up the same now. In the begining, it was just one client here and there (never thought anything serious about this), now, it is happening to ALL my clients... Not a single one of them can manage to stay connected for more than 5s (give or take a few).
[blha@blah] > interface wireless info print
0 interface-type=Prism firmware="1.1.0/1.7.4, compid 0x800c"
tx-power-control=no ack-timeout-control=no alignment-mode=yes
virtual-aps=no noise-floor-control=no scan-support=yes burst-support=no
supported-bands=2GHz-B
2GHz-B-channels=2412,2417,2422,2427,2432,2437,2442,2447,2452,2457,2462
This really sucks now... Hmmm