[kde-freebsd] Annoying network folder notifications

[Bengt Ahlgren]

Kris Moore <kris at pcbsd.org> writes:

> On 02/14/2012 09:35, Bengt Ahlgren wrote:
>> Hi!
>>
>> This is probably not the right list for this question, but as this might
>> be specific to the FreeBSD install...
>>
>> I just upgraded to KDE 4.7.4 from 4.5.x (don't remember exactly) running
>> on a 8.2-STABLE from Feb 6.  As a result, I started to get a lot of
>> annoying notifications from the device notifier about network folders on
>> the local network.  There are also a lot of TCP connections being setup,
>> or tried, currently for example I have *39* of these:
>>
>> tcp4       0      0 <mybox>.43716          <otherbox>.2869          TIME_WAIT
>>
>> And some of these:
>>
>> tcp4       0      0 <mybox>.54887          <otherbox2>.2869         ESTABLISHED
>> tcp4       0      0 <mybox>.54886          <otherbox2>.2869         ESTABLISHED
>> tcp4       0      0 <mybox>.30912          <otherbox>.2869          ESTABLISHED
>>
>> I think these connections are related to the notifications of the
>> network folders.
>>
>> Is this normal?  It is a bit excessive, I think.
>>
>> I have not found any KDE system setting to control this behaviour.  I am
>> usually uninterested of network folders, and would rather explicitly
>> browse the network when (if) needed.
>>
>> Bengt
>> _______________________________________________
>> kde-freebsd mailing list
>> kde-freebsd at kde.org
>> https://mail.kde.org/mailman/listinfo/kde-freebsd
>> See also http://freebsd.kde.org/ for latest information
>
> I too would be interested in a way to disable this. I think it may be
> related to an issue I'm seeing on a number of systems here with 4.7.3 as
> well, where several kdeinit4 processes constantly take up 100% CPU, and
> don't stop until I unplug the network cable, or "down" the interface.
> (It only happens depending upon the network I'm connected to)

On my system knotify4 is mostly lurking at about 1% CPU at all times,
but sometimes about four kdeinit4 processes go berserk and consume most
of the CPU, but this luckily happens seldom.

I tried to analyse this a bit with tcpdump.  It apparently is some UPNP
thing.  This is one such tcp connection on destination port 2869.  My
box transmits this:

SUBSCRIBE /upnphost/udhisapi.dll?event=uuid:yyyyyyyy-yyyy-yyyy-yyyy-yyyyyyyyyyyy+urn:upnp-org:serviceId:ContentDirectory HTTP/1.1
TIMEOUT: Second-1800
SID: uuid:xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
DATE: Tue, 14 Feb 2012 18:48:26
Connection: close
HOST: xx.yy.zz.ww:2869
content-length: 0

And gets this in response:

HTTP/1.1 200 OK
Server: Microsoft-Windows-NT/5.1 UPnP/1.0 UPnP-Device-Host/1.0 Microsoft-HTTPAPI/2.0
Timeout: Second-300
SID: uuid:xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx
Date: Tue, 14 Feb 2012 17:47:52 GMT
Connection: close
Content-Length: 0

xx.yy.zz.ww is the IP of the other box.  The SID uuid:s are the same,
but different from the one in the SUBSCRIBE.

Bengt