Karlnet
[Top] [All Lists]

Re: [Karlnet] AP1000 keeps loosing one card

To: karlnet@WISPNotes.com
Subject: Re: [Karlnet] AP1000 keeps loosing one card
From: Chris Conn <cconn@abacom.com>
Reply-to: karlnet@WISPNotes.com
Date: Thu, 05 Dec 2002 12:54:31 -0500
List-post: <mailto:karlnet@WISPNotes.com>
helio@compuland.com.br wrote:
Hi:

 Unfortunately turning off watchdog reboot timer on the base station do not
solve the lockup issue. I've done that a week ago. Today, 10 minutes ago,
one base locked one of it's cards again. Luckily I was able to connect to the
bridge remotely using the other card in the bridge and a remote reboot solved.

What Karlnet say about that issue ? Does not seem to be a isolated
problem...

I'm running Karlnet ISP BASE 4.02 in a brand new AP1000 with two
Orinoco gold Cards.

Hello,


Would you mind checking it when it is down/locked-up under "Interface Monitor"; choose the dead card and look to see if the "Operational Status" is UP or DOWN. if it is listed as DOWN, I have seen this problem before, but only with AVAYA radios. Putting a Orinocco silver instead (with 8.10 firmware, even if the 4.x code seems to want to downgrade them...) seems to have solved the problem. We swapped 3 AVAYA radios before using an old dusty Orinocco. I am told that Orinocco and Avaya cards are different as they use a different driver, however the FCC ID and markings are quite similar to not say identical.

If it is up, can you check to see if you still have "Turbocell Station Entries" for that card on the base? Are there continual retransmits but no packets to these satellites? If so, it is another problem we came across on many occasions and we believe by using a PRISM2.5 card we have solved it. There are some list readers that can hook you up with one. They are higher output as well so you can increase your signal considerably. Too bad Karlnet hasn't given those PRISM cards the ability to report proper signal information so the wireless link tests look aweful...

Chris


<Prev in Thread] Current Thread [Next in Thread>