Connecting laptop crashes network!!

Last week I was sent to a job which was outside of my usual remit (I
normally work on printers).

Anyway, the details I had were very sketchy to say the least....."user
requires advice on laptop, was sent a re-build disk but doesn't know what to
do with it" That was all I had to go on!

Spoke to user who wasn't on site and got a bit more info. She said that the
laptop had had connectivity problems (apparently when she connected it to
the network in Manchester, it brought all the company's network down!) and
had been re-built from a disk image. I was merely required to log it on to
the company network and test it.

She told me the training passwords but none of these seemed to work so I
asked a user to log in as they use roaming profiles. At the same time as
this, I could hear various members of staff.... "I can't get into my
e-mails"......"My systems' frozen"...."I've lost my connection, the server
must have gone down" etc.etc.

Without saying a word, I unplugged the network cable and, as I watched a
user's screen, her Outlook suddenly sprang to life!!....."My e-mails are
back"....."the systems' back up" etc.

For the life of me, I cannot see how a single PC/laptop can crash a whole
network - especially when it's been wiped & restored to original settings,
ruling out possible virus scenarios.

Any ideas??

TIA

Nel wrote:
> Last week I was sent to a job which was outside of my usual remit (I
> normally work on printers).
>
> Anyway, the details I had were very sketchy to say the least....."user
> requires advice on laptop, was sent a re-build disk but doesn't know
> what to do with it" That was all I had to go on!
>
> Spoke to user who wasn't on site and got a bit more info. She said
> that the laptop had had connectivity problems (apparently when she
> connected it to the network in Manchester, it brought all the
> company's network down!) and had been re-built from a disk image. I
> was merely required to log it on to the company network and test it.
>
> She told me the training passwords but none of these seemed to work
> so I asked a user to log in as they use roaming profiles. At the
> same time as this, I could hear various members of staff.... "I can't
> get into my e-mails"......"My systems' frozen"...."I've lost my
> connection, the server must have gone down" etc.etc.
>
> Without saying a word, I unplugged the network cable and, as I
> watched a user's screen, her Outlook suddenly sprang to
> life!!....."My e-mails are back"....."the systems' back up" etc.
>
> For the life of me, I cannot see how a single PC/laptop can crash a
> whole network - especially when it's been wiped & restored to
> original settings, ruling out possible virus scenarios.
>
> Any ideas??

One possibility is that it has a static IP address or a dhcp reservation ,
same as the default gateway on the network , so PCs can't communicate with
any servers elsewhere.

Another one is that the network card is faulty and it's causing the switch
to lock up. If they have a spare switch that you can do some testing with ,
this could be easily proved or disproved.


--
Alex

"I laugh in the face of danger. Then I hide until it goes away"

www.drzoidberg.co.uk www.ebayfaq.co.uk

"Dr Zoidberg" <alexNOOOOOO!!!!!!!@drzoidberg.co.uk> wrote in message
news:(E-Mail Removed)...
> Nel wrote:
>> Last week I was sent to a job which was outside of my usual remit (I
>> normally work on printers).
>>
>> Anyway, the details I had were very sketchy to say the least....."user
>> requires advice on laptop, was sent a re-build disk but doesn't know
>> what to do with it" That was all I had to go on!
>>
>> Spoke to user who wasn't on site and got a bit more info. She said
>> that the laptop had had connectivity problems (apparently when she
>> connected it to the network in Manchester, it brought all the
>> company's network down!) and had been re-built from a disk image. I
>> was merely required to log it on to the company network and test it.
>>
>> She told me the training passwords but none of these seemed to work
>> so I asked a user to log in as they use roaming profiles. At the
>> same time as this, I could hear various members of staff.... "I can't
>> get into my e-mails"......"My systems' frozen"...."I've lost my
>> connection, the server must have gone down" etc.etc.
>>
>> Without saying a word, I unplugged the network cable and, as I
>> watched a user's screen, her Outlook suddenly sprang to
>> life!!....."My e-mails are back"....."the systems' back up" etc.
>>
>> For the life of me, I cannot see how a single PC/laptop can crash a
>> whole network - especially when it's been wiped & restored to
>> original settings, ruling out possible virus scenarios.
>>
>> Any ideas??
>
> One possibility is that it has a static IP address or a dhcp reservation ,
> same as the default gateway on the network , so PCs can't communicate with
> any servers elsewhere.
>
> Another one is that the network card is faulty and it's causing the switch
> to lock up. If they have a spare switch that you can do some testing with
> , this could be easily proved or disproved.

I would have thought more likely to be hardware than software (i.e. IP
address) as they use a standard hard disk build.

Mind, having said that, I did attend another laptop call in Harrogate a few
weeks ago that certainly was a glich on the software side. Laptop had been
replaced twice and re-built about 5 times. Would connect to network when
plugged directly into laptop, but not when connected to the network card of
the docking station. I was asked to replace (again!), build and test.
Before going down the replace route, I had a play around first and found
that (IIRC) the TCP/IP setting was not enabled. Got the user up and running
in around 15mins as opposed to virtually all day replacing, re-building,
downloading software & testing!! Even had time to call into Leeds and get
me dad some tripe & brawn!! :-)

Anyway, the laptop that kills networks is no longer 'owt' to do with me as
the fella that normally does that contract was back the next day. No idea
what the outcome was but I just found it bizarre that such a small link
could disable the whole chain.

On 20 Feb, 23:03, "Nel" <nel...@despammed.com> wrote:
> "Dr Zoidberg" <alexNOOOOOO!!!!...@drzoidberg.co.uk> wrote in message
>
> news:(E-Mail Removed)...
>
>
>
>
>
> > Nel wrote:
> >> Last week I was sent to a job which was outside of my usual remit (I
> >> normally work on printers).
>
> >> Anyway, the details I had were very sketchy to say the least....."user
> >> requires advice on laptop, was sent a re-build disk but doesn't know
> >> what to do with it" That was all I had to go on!
>
> >> Spoke to user who wasn't on site and got a bit more info. She said
> >> that the laptop had had connectivity problems (apparently when she
> >> connected it to the network in Manchester, it brought all the
> >> company's network down!) and had been re-built from a disk image. I
> >> was merely required to log it on to the company network and test it.
>
> >> She told me the training passwords but none of these seemed to work
> >> so I asked a user to log in as they use roaming profiles. At the
> >> same time as this, I could hear various members of staff.... "I can't
> >> get into my e-mails"......"My systems' frozen"...."I've lost my
> >> connection, the server must have gone down" etc.etc.
>
> >> Without saying a word, I unplugged the network cable and, as I
> >> watched a user's screen, her Outlook suddenly sprang to
> >> life!!....."My e-mails are back"....."the systems' back up" etc.
>
> >> For the life of me, I cannot see how a single PC/laptop can crash a
> >> whole network - especially when it's been wiped & restored to
> >> original settings, ruling out possible virus scenarios.
>
> >> Any ideas??
>
> > One possibility is that it has a static IP address or a dhcp reservation ,
> > same as the default gateway on the network , so PCs can't communicate with
> > any servers elsewhere.
>
> > Another one is that the network card is faulty and it's causing the switch
> > to lock up. If they have a spare switch that you can do some testing with
> > , this could be easily proved or disproved.
>
> I would have thought more likely to be hardware than software (i.e. IP
> address) as they use a standard hard disk build.
>
> Mind, having said that, I did attend another laptop call in Harrogate a few
> weeks ago that certainly was a glich on the software side. Laptop had been
> replaced twice and re-built about 5 times. Would connect to network when
> plugged directly into laptop, but not when connected to the network card of
> the docking station. I was asked to replace (again!), build and test.
> Before going down the replace route, I had a play around first and found
> that (IIRC) the TCP/IP setting was not enabled. Got the user up and running
> in around 15mins as opposed to virtually all day replacing, re-building,
> downloading software & testing!! Even had time to call into Leeds and get
> me dad some tripe & brawn!! :-)
>
> Anyway, the laptop that kills networks is no longer 'owt' to do with me as
> the fella that normally does that contract was back the next day. No idea
> what the outcome was but I just found it bizarre that such a small link
> could disable the whole chain.- Hide quoted text -
>
> - Show quoted text -

<points up>
as dr z pointed out, it'll be an IP clash or possibly a faulty NIC.
best one i've seen along these lines here involved a dodgy laptop
dock. We first noticed that one satellite office had completely
vanished off the network. this was in the days before we used VLANs
so troubleshooting took *some time*. we eventually, though a process
of elimination, tracked it to a specific range of 8 ports. wandering
from office to office, we looked for plugged-in kit. after a
fruitless afternoon, we eventually tracked it down to the laptop in
question, which was connected to a docking station and the whole lot
locked in a desk drawer.
after pulling the plug on it everything returned to normal.
we eventually discovered that the docking bars for toshiba 3480s and
3420s, whilst they looked identical, had one crucial difference. use
the wrong one, and the n/w card would broadcast crap that'd saturate
and kill the network...wonderful design from toshiba...

ric

"ric" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed) oups.com...
> On 20 Feb, 23:03, "Nel" <nel...@despammed.com> wrote:
>> "Dr Zoidberg" <alexNOOOOOO!!!!...@drzoidberg.co.uk> wrote in message
>>
> we eventually tracked it down to the laptop in
> question, which was connected to a docking station and the whole lot
> locked in a desk drawer.
> after pulling the plug on it everything returned to normal.
> we eventually discovered that the docking bars for toshiba 3480s and
> 3420s, whilst they looked identical, had one crucial difference. use
> the wrong one, and the n/w card would broadcast crap that'd saturate
> and kill the network...wonderful design from toshiba...

LOL. Toshiba should be roasted slowly over a spit for such a stupid design.

It's almost as good as the design fault that allows an RJ11 phone lead to be
plugged into an RJ45 Ethernet socket instead of a dial-up modem: the
narrower plug will fit into the wider socket, but if you do it by accident
(I had a customer do it the other day and I know several other people who've
done it absent-mindedly) then the phone line gets shorted out and there's no
dialling tone on any phone on that line; I'm not sure what effect it has at
the exchange but I can imagine BT having kittens when they find a shorted
line.

In message <45dc5ea0$0$8730$(E-Mail Removed)>, Martin
Underwood <(E-Mail Removed)> writes
>LOL. Toshiba should be roasted slowly over a spit for such a stupid design.
>
>It's almost as good as the design fault that allows an RJ11 phone lead to be
>plugged into an RJ45 Ethernet socket instead of a dial-up modem:
Actually, I don't think it's a design fault, ISTR reading that they were
designed that way for some arcane purpose!
>the
>narrower plug will fit into the wider socket, but if you do it by accident
>(I had a customer do it the other day and I know several other people who've
>done it absent-mindedly)
Allegedly, the late great Douglas Adams also fell foul of this when
trying to connect a Mac to the Internet.

--
Clint Sharp

Martin Underwood wrote:

> It's almost as good as the design fault that allows an RJ11 phone lead to be
> plugged into an RJ45 Ethernet socket instead of a dial-up modem: the
> narrower plug will fit into the wider socket, but if you do it by accident
> (I had a customer do it the other day and I know several other people who've
> done it absent-mindedly) then the phone line gets shorted out and there's no
> dialling tone on any phone on that line; I'm not sure what effect it has at
> the exchange but I can imagine BT having kittens when they find a shorted
> line.
>
>
http://en.wikipedia.org/wiki/RJ45
"in floodwired environments the center (blue) pair is often used to
carry telephony signals. Where so wired, the physical layout of the 8P8C
modular jack allows for the insertion of an RJ11 plug in the center of
the socket, provided the RJ11 plug is wired in true compliance with the
U.S. telephony standards (RJ11) using the center pair."

"Clint Sharp" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> In message <45dc5ea0$0$8730$(E-Mail Removed)>, Martin
> Underwood <(E-Mail Removed)> writes
>>LOL. Toshiba should be roasted slowly over a spit for such a stupid
>>design.
>>
>>It's almost as good as the design fault that allows an RJ11 phone lead to
>>be
>>plugged into an RJ45 Ethernet socket instead of a dial-up modem:
> Actually, I don't think it's a design fault, ISTR reading that they were
> designed that way for some arcane purpose!

Irrespective of whether RJ11s were designed to be plugged into RJ45s, the
choice of those two sockets is a design issue in that it fails the "what are
the implications of someone accidentally doing it" idiot-proofing test.
Given that the RJ11 has probably been around for longer as the US telephone
socket than the RJ45 has been as the Ethernet socket (before twisted pair
everyone used BNC or D connectors and external tranceivers), the fault
probably lies with whoever specified the RJ45 for Ethernet.

>>the
>>narrower plug will fit into the wider socket, but if you do it by accident
>>(I had a customer do it the other day and I know several other people
>>who've
>>done it absent-mindedly)
> Allegedly, the late great Douglas Adams also fell foul of this when trying
> to connect a Mac to the Internet.

So it's not just me, my dad, my customer and several other people who have
done it. When the Ethernet is on a PCi card right next to the modem, it's
easy to get it wrong when you are feeling round the back of a PC in poor
light under a desk.

In message <45dc9cd1$0$8735$(E-Mail Removed)>, Martin
Underwood <(E-Mail Removed)> writes
>So it's not just me, my dad, my customer and several other people who have
>done it. When the Ethernet is on a PCi card right next to the modem, it's
>easy to get it wrong when you are feeling round the back of a PC in poor
>light under a desk.
Ahh, I have a perfectly sized little finger for that, if it fits into
the socket then it's Ethernet, if it doesn't then it's a modem.
>
>

--
Clint Sharp

"Kev" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
> Martin Underwood wrote:
>
>> It's almost as good as the design fault that allows an RJ11 phone lead to
>> be plugged into an RJ45 Ethernet socket instead of a dial-up modem: the
>> narrower plug will fit into the wider socket, but if you do it by
>> accident (I had a customer do it the other day and I know several other
>> people who've done it absent-mindedly) then the phone line gets shorted
>> out and there's no dialling tone on any phone on that line; I'm not sure
>> what effect it has at the exchange but I can imagine BT having kittens
>> when they find a shorted line.
> http://en.wikipedia.org/wiki/RJ45
> "in floodwired environments the center (blue) pair is often used to carry
> telephony signals. Where so wired, the physical layout of the 8P8C modular
> jack allows for the insertion of an RJ11 plug in the center of the socket,
> provided the RJ11 plug is wired in true compliance with the U.S. telephony
> standards (RJ11) using the center pair."

Sounds fair enough. I presume the centre two pins are not used by Ethernet.
So maybe British modem phone leads are not wired in the same way so as to
use the centre two pins. I wonder why not. I presume it means that modems
for use in the UK usually are wired in the same non-US way. Is this because
in Britain we have the resistor and capacitor in the socket whereas
elsewhere they are in the phone/modem? So maybe the shorting thing is only a
problem in Britain and not the whole world.

By the way, I didn't mention the 8P8C versus RJ45 issue: apparently the
so-called RJ45 plug/socket used for Ethernet is correctly called 8P8C (8
pins, 8 connectors); a true RJ45 as used in the telecomms industry has
either an asymmetric plug or else a lug on the side.