You are not logged in. Login Now
 0-24   25-49   29-53   54-78   79-86      
 
Author Message
25 new of 86 responses total.
gull
response 54 of 86: Mark Unseen   Apr 16 21:28 UTC 2005

The logging level probably sets how verbose the log is.  Usually with these
sorts of things a setting of 0 will log only critical errors, and a log
level of 7 logs a huge quantity of debugging information.  Often the most
useful setting is somewhere in the lower half of the range, but you'll have
to experiment.

I'm not familiar with OS X's logging, so I can't tell you what file to look
in.  You could temporarily turn up the logging level to something fairly
high, then search the log files for the base station's IP address.

I also don't know if OS X is set, by default, to accept log messages from
remote devices.  That feature is often turned off on other OS's, because it
can easily be used as an unauthenticated, remotely-accessable disk filling
service. ;)
scott
response 55 of 86: Mark Unseen   Apr 17 00:15 UTC 2005

OS X does text-based logging in (I think) /var/logs.  Basically the Unix
approach, and when I was first trying to figure out a modem problem on my
iBook it was easy to open a terminal window and find the logs.
rcurl
response 56 of 86: Mark Unseen   Apr 17 02:23 UTC 2005

I upped the Logging Level to "Informational" (6). In /var/log are many
files (and 6 directories). Those that are not empty are: CDIS.custom,
cups(dir), daily.out, system.log, windowserver.log, and
windowserver_last.log. I can't find an IP address (10.0.1.1 for the Base
Station) in any of them.

I directed the Base Station to log to this computer, which  has IP 
10.0.1.2. 
arthurp
response 57 of 86: Mark Unseen   Apr 18 06:17 UTC 2005

You'll probably need to activate a feature on 10.0.1.2 to allow other
machines to write to it's logs.  'man syslogd' might have the info you
need.  'apropos syslog' may be a longer route to the info.  Once it is
ready to receive log messages it will be listening on port 514 udp.
rcurl
response 58 of 86: Mark Unseen   Apr 18 17:21 UTC 2005

Do you know that to be a feature of OS X?  There is a syslogd command, but
I don't know enough to set it up to log from the Base Station (if that can
even be done).

arthurp
response 59 of 86: Mark Unseen   Apr 21 01:10 UTC 2005

Your syslogd should have inherited the feature from its UNIX roots. 
Refer to the included manual page for syslogd from a Linux system.

    -r  This option will enable the facility to receive message from the
        network using an internet domain socket with the syslog  service
        (see  services(5)).   The default is to not receive any messages
        from the network.

        This option is introduced in version 1.3 of the  sysklogd  pack-
        age.   Please  note that the default behavior is the opposite of
        how older versions behave, so you might have to turn this on.

I don't know for sure how OS X sets options for services.  Linux would
use /etc/sysconfig/syslogd as a text file containing startup options for
syslogd.  It may be the same in OS X.  I think they use similar rc scripts.
rcurl
response 60 of 86: Mark Unseen   Jun 22 16:50 UTC 2005

I  would appreciate some information about DHCP in IP/TCP. In the Network
dialog in OS X I can implement automatic DHCP, but have no other control 
over it except to manually renew the DHCP lease. However in the Airport 
Base Station I'm using I can choose the DHCP lease term from minutes to 
days. What are the consequences of choosing a lease term? What happens
at the time a lease "expires" (and does its expiration depend upon whether 
the computer, or Airport Base, is in use, idle, or off)? 

The default lease term on the Airport Base was 4 hours, and on some
occasions I observed I no longer had internet access after my system had
been idle for more than 4 hours. Is this related, or a coincidence, or
what? If the former, how do I get a new lease immediately?

Mac Help is pretty useless in regard to these questions. 
scott
response 61 of 86: Mark Unseen   Jun 22 20:56 UTC 2005

The DHCP lease is something that the DHCP server sets, not the client.  For
your application I'd set the lease either off or for many days - a shorter
lease is only useful if you have other clients connecting and needing a lease
from a limited pool of IP addresses.  
rcurl
response 62 of 86: Mark Unseen   Jun 23 00:42 UTC 2005

Makes sense. Now what might I experience at lease expiration if I set the
term for several days (and how does this depend on whether the computers
and/or the Airport router is always on and/or used intermittently)? Or
maybe it is better to ask, what if any conditions of system operation and
lease term setting will lead to any observable consequence of setting a
finite lease term?

It doesn't look like I can set the lease term to zero.
gull
response 63 of 86: Mark Unseen   Jun 27 03:47 UTC 2005

A lease term of 0 wouldn't make any sense. 
 
The lease term sets how long a client gets to hold onto an IP address 
before it has to ask to have it renewed.  Setting the lease shorter 
means the IP will be available again sooner if the client goes away; 
setting it longer means that a client that isn't connected all the 
time will have a more stable IP address.  It doesn't generally affect 
how often the client asks to have the lease renewed, though.  My 
experience is that Windows machines, for example, will always ask for 
a renewal once every five minutes as long as they're online. 
 
rcurl
response 64 of 86: Mark Unseen   Jun 27 06:24 UTC 2005

I'm trying to understand how this applies in my system, where I have two
computers on a wired LAN that connects to a wireless router. I set the
DHCP lease term on the *router*, which is on most of the time.  The
computers are shut down often. When the lease term was set to 4 hours, on
occasion I could not connect from the computers until I had shut
everything down and restarted. Can this behavior be explained with how the
lease system works?
scott
response 65 of 86: Mark Unseen   Jun 27 12:01 UTC 2005

I don't think you should be losing connection that way.  
gull
response 66 of 86: Mark Unseen   Jun 27 13:46 UTC 2005

I don't think that's related, unless there's a bug somewhere.

All a lease term of 4 hours means is if the router doesn't hear from a
particular computer for 4 hours, it releases that machine's IP address
and may hand it out to another machine if a new request comes in.  The
original computer will still get an IP address when it asks for one
again, it just may not be the one it had originally.
rcurl
response 67 of 86: Mark Unseen   Jun 27 16:56 UTC 2005

Here's what I get for the IP addresses of the various devices:

iMac        G4        Bridge        Router
10.0.1.2    10.0.1.3  192.168.30.1  10.0.1.1
                                    69.137.222.90

I think that Bridge IP address is for accessing it's software to configure
it (?). The second one for the router is its "Public (WAN) IP address". 
The router has a default setting of
 
   "Share a single IP address (using DHCP and NAT)"

with an alternative, not being used of

   "Share a range of IP addresses (using only DHCP)"

Does this tell anyone anything about how this system is working (and if
it might lead to a disconnection if the lease term in the router is 4 
hours)?

gull
response 68 of 86: Mark Unseen   Jun 27 19:44 UTC 2005

It looks all right to me.  It's a little odd that the bridge's address
isn't in the same subnet, but for bridging purposes that really
shouldn't matter.
rcurl
response 69 of 86: Mark Unseen   Jun 30 05:47 UTC 2005

How long does it take for the router to get a new IP address at the end of a
lease term? If this can be a significant time, then I would have an increased
probability of being able to connect anew while it is negotiating for a new
lease if the lease term is short (like it was, at 4 hours). I have increased it
to 4 days, and I have not encountered the inability to connect since.
gull
response 70 of 86: Mark Unseen   Jun 30 14:10 UTC 2005

Usually it doesn't take more than a second or two.
rcurl
response 71 of 86: Mark Unseen   Jun 30 15:40 UTC 2005

Is it possible for the ISP (Comcast, in this case) to run out of leases? 
juicy
response 72 of 86: Mark Unseen   Jul 1 01:49 UTC 2005

I get the impression that, with Comcast, *any*thing is possible. . . . But
it should be improbable.
arthurp
response 73 of 86: Mark Unseen   Jul 16 15:18 UTC 2005

Given the 'always on' nature of cable internet, I would call it
unethical to sell more accounts than they have IPs.  So in correct
operation it should not be possible for them to run out of leases.  But
as juicy says...  I've had to debug and solve problems for them on a
couple occasions.
rcurl
response 74 of 86: Mark Unseen   Jan 23 21:02 UTC 2008

I have a wired Ethernet LAN with two computers (iMac and G4) and a printer 
(Brother 2070N), connected with a wireless adapter to an Airport base 
station (and from that to a cable router). I can also connect to the 
internet through the base station from a Mac laptop. I would like to 
*print* with the printer on the LAN from the laptop.

Is this possible and, if so, how? 

The Airport Extreme manuals only have information about printing to a USB
printer connected to the base station.
nharmon
response 75 of 86: Mark Unseen   Jan 24 00:19 UTC 2008

Assuming the airport base station is like other wireless router/access
points, the laptop should already be on the same network as the
workstations and printer.

Have you tried pinging the printer from the laptop?
rcurl
response 76 of 86: Mark Unseen   Jan 24 01:58 UTC 2008

From termimal? Using its DHCP IP address? I thought this was all built into
OSX. The Printer Browser on the laptop shows the printer, but gives no
location for it. (I'm in the dummy class for a lot of these networking details
so I need help.)
nharmon
response 77 of 86: Mark Unseen   Jan 24 04:37 UTC 2008

I would think you should be able to access the printer like you would
from the PCs. What kind of location information do the PCs give?
rcurl
response 78 of 86: Mark Unseen   Jan 24 04:58 UTC 2008

I got it to work with the DHCP IP address. The Mac Help helped but the 
procedure is not intuitive. What led me astray is that when I installed 
the printer software, it showed up in the printer selection menu, which I 
thought meant that it saw the printer. But no: one has still to Add the 
printer and give it an arbitrary name and location. There is then an 
obscure menu, which is hard to find, where one chooses the make and model 
of the printer (there are hundreds listed!).


 0-24   25-49   29-53   54-78   79-86      
Response Not Possible: You are Not Logged In
 

- Backtalk version 1.3.30 - Copyright 1996-2006, Jan Wolter and Steve Weiss