[Dxspider-support] FYI: 4 new nodes added to ROS Auto-Spottingfunction.

Bela Markus ha5di at freemail.hu
Thu Jul 15 12:31:49 BST 2010


  It looks like a Cold War starting. Criple mode seems to be a solution 
for a while, but for sure they will introduce a new version with 
different behavior to bypass such filters.

Core issue is that the whole DXCLUSTER system is an open system (open 
mean security wise) without real user authentication and control, and 
you can do whatever you can in SPIDER one can always find a hole in the 
system. I'm not saying do nothing specially as most nodes in the system 
are SPIDER's, but pessimistic and expecting a long war without final 
victory.

Regards... Béla, HA5DI



2010.07.15. 13:21 keltezéssel, Anthony J. Cioffi (N2KI) írta:
> I noticed two of those nodes connect to me for spots!
>
> BG2RVL-9
> BA2IA-2
>
> Suggestions?
>
> Anthony
>
>
> -----Original Message-----
> From: dxspider-support-bounces at dxcluster.org [mailto:dxspider-support-bounces at dxcluster.org] On Behalf Of SM6U, Rickard
> Sent: Thursday, July 15, 2010 5:14 AM
> To: The DXSpider Support list
> Cc: vk3ama.laurie at gmail.com; The DXSpider Support list
> Subject: Re: [Dxspider-support] FYI: 4 new nodes added to ROS Auto-Spottingfunction.
>
> I vote YES for cripple mode! Please make it configurable, but on by default.
>
> Even if it only covers thoose running updated spiders, it will sure cause trouble for the softwares that keep abusing the
> cluster network, possibly convincing the creators to change the code.
>
> /Rick, sm6u
>   on the phone
>
> 14 jul 2010 kl. 09:39 skrev Dirk Koopman<djk at tobit.co.uk>:
>
>> On 12/07/10 21:51, VK3AMA wrote:
>>> FYI
>>>
>>> The latest release of ROS software has added 4 new nodes to
>>> auto-spotting function.
>>>
>>> BG2RVL-9 dx.5iyou.cn:7373
>>> BD5RV dxc.jsdxc.net:7300
>>> BA2IA-2 spot.ba2ia.com:7300
>>> 9H1LO-1 dx.9h1lo.net:7300
>>>
>>> At time of this email, of the last 150 ROS cluster spots,
>>> 148 were auto-generated (98.7%).
>>>
>> If all major nodes were to do a "set/badword ros" then it probably does not matter that he keeps adding to the list, as the
> spots won't propagate very far.
>> On an other tack: I have been wondering for a while what to do with all the programs (of which the ROS stuff is but one)
> that insist on logging in, doing one or two things (eg send a spot or sh/dx), then immediately logging out.
>> Now, the code already does not bother (in most cases) to send out routing information for these in/out merchants, but I am
> now thinking that I ought perhaps to be a bit more brutal.
>> It would be a fairly simple job to have some kind of averaging or pump up timer that has the effect going into "cripple"
> mode if the last n connections were less than m seconds long.
>> "Cripple" mode would something like pretending to be a nice cooperative cluster that accepts spots, but does not send them
> on, and/or returns a *very* limited view of data such as spots (maybe only two or three spots else spots that are an hour or
> so old).
>> If the user starts to behave like a real person and connects p times for more than m seconds, it automatically comes out of
> "cripple" mode. While in "cripple" mode for more than (say) 2*m+random amount seconds then de-cripples that user for that
> session. This behaviour persists until they have behaved for p sessions.
>> "Misbehaviour" puts the user back into "cripple" mode.
>>
>> Views anyone?
>>
>> Dirk G1TLH
>>
>>
>> _______________________________________________
>> Dxspider-support mailing list
>> Dxspider-support at dxcluster.org
>> http://mailman.tobit.co.uk/mailman/listinfo/dxspider-support
> _______________________________________________
> Dxspider-support mailing list
> Dxspider-support at dxcluster.org
> http://mailman.tobit.co.uk/mailman/listinfo/dxspider-support
>
>
> _______________________________________________
> Dxspider-support mailing list
> Dxspider-support at dxcluster.org
> http://mailman.tobit.co.uk/mailman/listinfo/dxspider-support




More information about the Dxspider-support mailing list