Ben Nisenbaum wrote:
Hello readers,
I have a problem with my dial-up connection when my system is allocated the remote address of 210.8.1.13. When this address is given to me, I am unable to access the mail server and the internet, with one exception, the log of which is below. The mail programs I have used are Mozilla mail, Netscape mail and Thunderbird. Currently I use Fedora Core 3, but the same problem has existed for over the last few years using Red Hat 7.3, 9 and Fedora Core 1 and 2. When the failure to get through occurs, the mailer or browser stops at the "resolving address" point and then times out. Over the past year and a bit I have had numerous phone conversations and correspondence by letter with AAPT and its help desk and beyond. Originally the help desk people spoke about the distance of my line from the servers, and other things about my modem settings which all made little sense because on the 20 or so other remote addresses which I had been allocated over the years, all have worked perfectly well. In the end, at the end of last year, I spoke to someone with a little more seniority and he spoke to someone at Connect with my problem and came back to me with something like the following: "your machine is ripping the address of the access router itself, which is one of twenty at Burwood (Vicoria) ... your computer is grabbing number 3 and it shouldn't."
Not being a network mavin, I wasn't sure what this meant. Was he referring to the Routing Information Protocal when he said "ripping?" I couldn't do anything immediately with that information anyway. I conjectured that this router, 210.8.1.13, may have been configured differently to the other routers used by AAPT, but the reply to this suggestion was that the windows users had no trouble with the router and that therefore there was no reason for AAPT to consider even looking into it ... the linux people were on there own, they said. They suggested I post the issue on the slug list because slug had people who could help me. So here I am.
The solution to this problem, it appears to me, is to get their server to exclude
ip-number 210.8.1.13 from being assigned to any customer. If this number
is excluded then you as a dial-up customer will not be assigned this ip-number
ever. You as a customer has no control on this only your ISP.
The server software that does this assignment is one of several, namely: TACACS,
RADIUS, variations of the previous two, etc., depending on what your ISP is using.
With this software, the ip-numbers assignment may be controlled by SysAdmin
by excluding certain numbers. The ISP should do this because 210.8.1.13 is
already assigned as a permanent address to one of their routers, as I understand
from above. What is the idea of putting 210.8.1.13 in their pool of available
ip-addresses when it is assigned already ? Senseless, I'd say.
Incorrect subnetting may also cause this problem even if the given ip-address
is not used by another computer or router. If you wish to understand how
subnetting may cause this problem, there are several tutorials on subnetting from
the Internet. Again if subnetting is a problem your ISP has control and not you.
The onus is on your ISP to fix the problem. I like to refer to this link time and
again to refresh my understanding of subnetting:
http://www.cisco.com/warp/public/701/3.html
How is it possible for Windows dialler not to be assigned this ip-number ? If your Linux
is assigned this number it is almost certain that Windows is assigned this number
once every so often. Perhaps, the Windows user do not realise it. They just disconnect
and re-dial costing them some more $0.CENTS everytime and get comfort in the
idea that it is their telephone line that's a problem.
There is a work-around under this circumstance. Check if there is another local-call
point-of-presence to you and use that number instead of the current one. You might
get lucky.
The other ultimate solution is change ISP provider. But do not change until you have
contacted:
http://www.tio.com.au
to explain your problem
DISCLAIMER: Any statement in this post is not to be taken as a recommendation
and should you decide to take action as a consequence, you are doing it at your
own risk and perile. You have been warned.
The dial-up program that I have always used has been the standard /sbin/ifup ppp0 which uses wvdial. On occasion I have used wvdial itself to connect, but that has made no difference to my problem. I have tried to inform myself about the dial-up processes and protocols, and have inspected the logs which I have reproduced below. My untrained eye cannot detect any hint of the problem in these logs. I have read the relevant sections in "Using and Managing PPP" by Andrew Sun to understand the logs, and the troubleshooting section doesn't really refer to the sort of problem I have. All the relevant controls appear to be with the ISP machinery. Googling has not helped me with my specific concern.
In trying to encourage AAPT to look at the problem a little more deeply, I mentioned that it may be in their interest to do so in order to discover a solution that may benefit others. Unfortunately, in the end, I mentioned to them that the odd 16 or so cents for a phone call that every now and then resulted in hitting this router, should really be considered a needless loss to me. The next phone conversation after that was very frosty, and they said they were closing the file on this problem and that I was on my own with it. My expression of concern about the problem, and not the trifling cost, was disregarded.
My question is: is there anything I can do at my end to have this router work as well as all the others do for me?
Ben
........................SNIPPED.......................................... -- SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/ Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html
