Help: Network Card Oddity
sam
that I use for my web server. Yesterday, my router stopped working
for my telephone and internet *only* when I had the linux server
connected via cable (from box to router). When I unplugged it, and
plugged it back in, it worked for about an hour, then the phone
and internet went out again. If I leave the web server disconnected,
everything works fine.
Any thoughts?
Thanks,
s
Auric__
Either you're from the future, or else you mean Slackware 10.2. ;-)
--
Sometimes I lie awake at night and ask, "Why me?" Then a voice answers,
"Nothing personal, your name just happened to come up."
Lew Pitcher
Yah. Either your router or the ethernet card on your Linux system or
the cable between them is snafu.
The cheapest and easiest thing to fix is the ethernet UTP between the
Linux system and the router. Check this first. Reseat each end. If
that doesn't fix the problem, swap it out with a different cable.
The next cheapest/easiest thing to fix is the ethernet card in your
Linux system. Check to see that you are using the correct Linux
driver. Load the proper driver, if you need to. Take a look at /var/
log/syslog, /var/log/messages, and /var/log/debug to see what sort of
errors Linux is complaining about. Also look at your dmesg to see if
there are any kernel messages that might explain the problem. Swap out
your ethernet card with a new one (shouldn't cost more than $15 for a
good PCI ethernet card), load the proper driver, and check if that
fixes the problem.
Finally, the most costly thing to fix is the router. If it is
malfunctioning, you /might/ find a firmware update to fix it.
Otherwise, toss the thing out and replace it with a new router.
Other than those generic steps, there's nothing we can really tell
you.
HTH
Henrik Carlqvist
> Either your router or the ethernet card on your Linux system or
> the cable between them is snafu.
Another possibility is the power supply for the router. If it unable to
provide enough power too much load caused by connecting yet another
machine could cause the router to hang.
regards Henrik
--
The address in the header is only to prevent spam. My real address is:
hc3(at)poolhem.se Examples of addresses which go to spammers:
root@localhost postmaster@localhost
Dan C
> Lew Pitcher <lpit...@teksavvy.com> wrote:
>> Either your router or the ethernet card on your Linux system or the
>> cable between them is snafu.
>
> Another possibility is the power supply for the router. If it unable to
> provide enough power too much load caused by connecting yet another
> machine could cause the router to hang.
Ummmmmm...... Nope.
The power supply in a router doesn't power anything but the router.
Adding more connections to it does not affect the P/S at all.
--
"Ubuntu" -- an African word, meaning "Slackware is too hard for me".
"Bother!" said Pooh, as he garotted another passing Liberal.
Usenet Improvement Project: http://twovoyagers.com/improve-usenet.org/
goarilla
> On Fri, 06 Nov 2009 08:17:14 +0100, Henrik Carlqvist wrote:
>
>> Lew Pitcher <lpit...@teksavvy.com> wrote:
>>> Either your router or the ethernet card on your Linux system or the
>>> cable between them is snafu.
>>
>> Another possibility is the power supply for the router. If it unable to
>> provide enough power too much load caused by connecting yet another
>> machine could cause the router to hang.
>
> Ummmmmm...... Nope.
>
> The power supply in a router doesn't power anything but the router.
> Adding more connections to it does not affect the P/S at all.
it is possible though i had a crappy router that resetted when
i torrented and had more than 200 connections ( it
probably was just clearing the connection tables but still - i have
seen increased load situations that do kill the device )
then there's power over lan but i've never seen such device
so i assume they are very uncommon off the shelf
the thing is most home routers are flaky POS devices, if he has anything
decent maybe he could try to hook up a rs 232 cable to the router and
recreate the condition
sam
kevin....@skynet.remove-this.be says...
Yes, checked the cable, modem, and everything else worked
fine so long as I didn't hook up the Slackware box.
So...
I took the box to the garage and blew out the dust with my
air compressor. Then I did a fresh install of Slackware 13.
It's been a whole day and no problems so far.
I don't have great hopes, though. Thanks for the help.
s
goarilla
what about another box ? (mac os X, ubuntu, win^W)
i guess all the other boxes are always plugged in and
the slackware box is the new/black swan in the pond, it's the exotic
entity ?
> So...
>
> I took the box to the garage and blew out the dust with my air
> compressor. Then I did a fresh install of Slackware 13. It's been a
> whole day and no problems so far.
>
> I don't have great hopes, though. Thanks for the help.
>
> s
i understand,
you're sitting on pins and needles right now
Henrik Carlqvist
> The power supply in a router doesn't power anything but the router.
> Adding more connections to it does not affect the P/S at all.
You seem really sure that the number of connections and the activity on
those connections does not affect the power usage at all?
I have seen a few router which have leds indicating which network ports
are in use and if there is any activity on the ports. Knowing that leds
draw a little current when lit I feel rather sure that more connections
does affect the current the PS will have to supply. I would be rather
surprised if the router only needed current for the leds, my guess is that
the logic circuits in the router also needs current and I would also guess
that this current depends upon the load on the router.
Dan C
> Dan C <youmust...@lan.invalid> wrote:
>> The power supply in a router doesn't power anything but the router.
>> Adding more connections to it does not affect the P/S at all.
>
> You seem really sure that the number of connections and the activity on
> those connections does not affect the power usage at all?
>
> I have seen a few router which have leds indicating which network ports
> are in use and if there is any activity on the ports. Knowing that leds
> draw a little current when lit I feel rather sure that more connections
> does affect the current the PS will have to supply. I would be rather
> surprised if the router only needed current for the leds, my guess is
> that the logic circuits in the router also needs current and I would
> also guess that this current depends upon the load on the router.
The additional LED's make a negligible difference, and the "logic
circuits" are already powered up regardless of load.
So, yes, I'm sure that my original statement is correct.
Grant
>On Fri, 06 Nov 2009 20:40:54 +0100, Henrik Carlqvist wrote:
>
>> Dan C <youmust...@lan.invalid> wrote:
>>> The power supply in a router doesn't power anything but the router.
>>> Adding more connections to it does not affect the P/S at all.
>>
>> You seem really sure that the number of connections and the activity on
>> those connections does not affect the power usage at all?
>>
>> I have seen a few router which have leds indicating which network ports
>> are in use and if there is any activity on the ports. Knowing that leds
>> draw a little current when lit I feel rather sure that more connections
>> does affect the current the PS will have to supply. I would be rather
>> surprised if the router only needed current for the leds, my guess is
>> that the logic circuits in the router also needs current and I would
>> also guess that this current depends upon the load on the router.
>
>The additional LED's make a negligible difference, and the "logic
>circuits" are already powered up regardless of load.
Except for a certain netgear router with weak power supply that
did struggle when too many LEDs lit, active channels ;) Forget
the model, it was Gigabit 8 channel.
Grant.
--
http://bugsplatter.id.au
Henrik Carlqvist
> On Fri, 06 Nov 2009 20:40:54 +0100, Henrik Carlqvist wrote:
>> You seem really sure that the number of connections and the activity on
>> those connections does not affect the power usage at all?
> The additional LED's make a negligible difference,
Yes, but in swedish we have a saying which would translate to something
like "the final little drop that made the entire goblet to flood".
> and the "logic circuits" are already powered up regardless of load.
That might be rather true if the logic is built upon MOSFET technology,
however, with CMOS technology the power really depends upon usage. From
http://en.wikipedia.org/wiki/Cmos :
"Significant power is only drawn when the transistors in the CMOS device
are switching between on and off states."
> So, yes, I'm sure that my original statement is correct.
I would say, depending on chosen technology, you might not be right.
Dan C
> Dan C <youmust...@lan.invalid> wrote:
>> On Fri, 06 Nov 2009 20:40:54 +0100, Henrik Carlqvist wrote:
>>> You seem really sure that the number of connections and the activity
>>> on those connections does not affect the power usage at all?
>
>> The additional LED's make a negligible difference,
>
> Yes, but in swedish we have a saying which would translate to something
> like "the final little drop that made the entire goblet to flood".
>
>> and the "logic circuits" are already powered up regardless of load.
>
> That might be rather true if the logic is built upon MOSFET technology,
> however, with CMOS technology the power really depends upon usage. From
> http://en.wikipedia.org/wiki/Cmos :
>
> "Significant power is only drawn when the transistors in the CMOS device
> are switching between on and off states."
>
>> So, yes, I'm sure that my original statement is correct.
>
> I would say, depending on chosen technology, you might not be right.
I would say that means that I might be right.
james
@news.supernews.com:
> that I use for my web server. Yesterday, my router stopped working
> for my telephone and internet *only* when I had the linux server
I had a similar condition on a home box that I was using pretty
heavily from work. My local internet provider would notice the
heavy usage and cause my DHCP lease to expire.
Some of the software I was running wouldn't notice that, and
!poof!
I was also using dyndns.org. Updating it once an hour didn't
do much good when the address would change 10 minutes later.
--
The email address, above, is most certainly munged. Perhaps you
might reply to the newsgroup, instead? Thanks!
Jack Myers
> On Fri, 06 Nov 2009 20:40:54 +0100, Henrik Carlqvist wrote:
> > Dan C <youmust...@lan.invalid> wrote:
> >> The power supply in a router doesn't power anything but the router.
> >> Adding more connections to it does not affect the P/S at all.
> >
> > You seem really sure that the number of connections and the activity on
> > those connections does not affect the power usage at all?
> >
> The additional LED's make a negligible difference, and the "logic
> circuits" are already powered up regardless of load.
I've seen cases where the building cabling made a difference, so I'm sure
that the router is also driving some of the Ethernet signals, the complex
cable impedance, and any wiring problems as well as logic and LEDs.
Whenever I hear a problem like the OP has I want to look into equipment
grounding issues--lots of older houses have no grounded power outlets
and the chassis of the "different computers will have varying AC voltage
levels wrt "true ground.
If these computers have to fight it out via router ports over Ethernet
cables that are sometimes connected, sometimes not, the results can
be mildly amusing. Consider that many routers have auto-sense ports.
--
Jack Myers / Westminster, California, USA
In theory there's no difference between theory and practice;
in practice there is.
Chick Tower
> On Sat, 07 Nov 2009 09:07:46 +0100, Henrik Carlqvist wrote:
>
>> Dan C <youmust...@lan.invalid> wrote:
>>> So, yes, I'm sure that my original statement is correct.
>>
>> I would say, depending on chosen technology, you might not be right.
>
> I would say that means that I might be right.
Would you two please stop gazing at the other's navel?
--
Chick Tower
For e-mail: aols2 DOT sent DOT towerboy AT xoxy DOT net
Sylvain Robitaille
I can contribute to, at little risk of continuing to beat the already
dead horse ...
On 06 Nov 2009 17:22:13 GMT, goarilla wrote:
> ... i had a crappy router that resetted when i torrented and had more
> than 200 connections ...
Perhaps, but this is not a power-supply issue, which is what had been
suggested to this point.
> then there's power over lan ...
In a switch, not in a router.
Well, except that I'm not sufficiently familiar with the
"consumer-electronics" versions of these to know whether PoE
(power over Ethernet) has been built into any of the all-in-one
router/switch/wireless-access-point/broadband-gateway things ...
In "real" equipment, though, Power over Ethernet is provided from the
switch or from a separate device inline between the switch and the
powered device.
--
----------------------------------------------------------------------
Sylvain Robitaille s...@encs.concordia.ca
Systems analyst / AITS Concordia University
Faculty of Engineering and Computer Science Montreal, Quebec, Canada
----------------------------------------------------------------------
Sylvain Robitaille
> I took the box to the garage and blew out the dust with my
> air compressor. Then I did a fresh install of Slackware 13.
> It's been a whole day and no problems so far.
I've seen similar situations. If it acts up again, remove socketed
items (whether they be PCI cards, memory modules, CPUs and other chips),
and blow air into the sockets as well. Reassemble carefully, and the
system should be fine. Cleaning out dust is one of the first things I
do to computers (or other electronic items) that are acting up. It can
often be the only required remedy.
Grant
>On Fri, 6 Nov 2009 11:59:17 -0600, sam wrote:
>
>> I took the box to the garage and blew out the dust with my
>> air compressor. Then I did a fresh install of Slackware 13.
>> It's been a whole day and no problems so far.
>
>I've seen similar situations. If it acts up again, remove socketed
>items (whether they be PCI cards, memory modules, CPUs and other chips),
>and blow air into the sockets as well. Reassemble carefully, and the
>system should be fine. Cleaning out dust is one of the first things I
>do to computers (or other electronic items) that are acting up. It can
>often be the only required remedy.
And, check for bulging or leaking capacitors around the CPU socket and
graphics card, these should be replaced. Symptom of cap. fault is that
the mobo is reluctant to startup, but once started seems to work okay.
If left alone eventually the CPU and/or other chips blow up due to
power supply spiking.
I've seen bulging caps in the PSU too, it's worth opening up the PSU
to blow dust out and check the fan bearings oiled, as well as for bad
caps.
Grant.
--
http://bugsplatter.id.au
sam
g_r_a...@bugsplatter.id.au says...
It's been a week and it's still going. It might be fine for
a while, but I'm still scoping around on Craiglist for a cheap
backup, just in case.
s