[NoCat] OT: Can a bad cable cause interrupt problems?

[Brian Capouch]

Please forgive me for posting this to these groups, it is "wireless 
Linux" but neither NoCat nor HostAP.

I have an 802.11b setup using Linux laptops as CPE; I actually have 
about a dozen such installations.  Each station uses a Lucent silver to 
talk to an AP-1000, and each then distributes b/w via Ethernet to the 
client side.

I am at my wit's end with the following problem: one of my stations 
periodically goes down for periods of time running from a half-hour to 
several hours.

While it's down, the kernel logs fill with this message, over and over:

 > Nov 26 09:44:15  kernel: wvlan_cs: Initialization failed!
 > Nov 26 09:44:19  kernel: wvlan_cs: eth1 Tx timed out! Resetting card

Those same messages are reported by dmesg, except with a "NETDEV: 
watchdog timeout" type of message.

The real rub is this: this same thing happens despite my having switched 
  out COMPLETELY, all of the hardware up to the pigtail.  Three separate 
machines all exhibit this same behavior, and each works flawlessly when 
I remove it from this site.

So to me this means that somehow these errors are being induced by 
something in either the antenna proper or the transmission line.

Could it be bad grounding causing a voltage leak into the radio?  Could 
intermittent lack of connectivity do such a thing?

My apologies again for the OT post, and thanks in advance for any advice 
that might be forthcoming.

B.

p.s. I use wvlan_cs because orinoco_cs has some sort of weird 
incompatibility with something else in my hardware config. . . .


_______________________________________________
NoCat mailing list
NoCat at lists.nocat.net
http://lists.nocat.net/mailman/listinfo/nocat