Stopping NetBIOS Spam?

MoonFacedAssassin asks: "I woke up this morning to find that my computer had a Windows messaging pop-up window with an advertisement about getting diplomas and degrees. I was quite shocked to find that my Bellsouth DSL IP address had been spammed. Has this happened to anyone else? Other than closing off the port which this can come through, are there any other ways to block this spam? And, how responsible is Bellsouth (or any ISP for that matter) in handling issues like this?"

Comments

Firewall yourself...

[earthdark]

TechTV covered this earlier this year so you might want to read their breif article for more information.

Basically, they're port scanning for open port 139s and spam IP that comes up positive. Either turn off the messenger service in services or install a firewall/router and block incoming tcp connections on port 139 (NetBIOS).

While you're at it, turn off the remote registry service...

Re:um.

[pruneau]

I concur with you danielrose.

<RANT> Without denying MoonFacedAssassin the right to a response to a very pertinent question, I think that posting that on Ask SlashJeeves shows a suprising level of ignorance from Cliff. Or he is at the end of some coding spree? Because this question does dot belong here, but rather on some newsgroup like comp.security.firewall (someone help me there).

Anyway, let me end my RANT section by saying that the level of interest of ask slashdot has regularly reached new lows every day.

I'm worried, to say the least

</RANT>

But let's drop the political/marketing aspect of that and take car of some real technical stuff:

(Yeah you guessed it, I'm getting pedantic during insomnias (it's 3:00 here)).

1. Get over it: an IP does not get spammed, because scannig block of addresses does not require any kind of disclosure from your ISP. They only have to have a router advertising their block of IP to the internet for those block of addresses to be scanned. Because having such a setup is one of the primary requirements to be an ISP. Sorry. Even residing into some secret whois database won't change anything there.

   Believe me, I've got firsthand experiences of having systems simply plug onto the internet, not even having some DNS record, and beeing scanned after one days of routable IP presence.

   And no, you don't want them to "protect" you from that, because if they start going big brother on you, you will notice a real drop on the number of things you can do online. Unless you really want only to surf and e-mail a bit, that will be perfectly understandable.

   But that's another debate: since internet is a jungle now, do we want to see some new kind of ISP that babysits theyr not-so-technical users ?

2. The second thing that worries me is that if you got a window messaging message, this means that you probably have a whole slew of netbios services exposed to the internet. Now listen carefully: if you ever have a shared printer or worst, drive, your machine is already hacked. Even not having shares might no be sufficient to protect you.

   I'm not trying to scare you there, it`s just a fact.

   In this case, please unplug from the network and reinstall from scratch. Do not backup any executablte. And the first time you re-plug you machine on the internet, please go immediately shopping for a personnal firewall, like ZoneAlarm and such. Once this one is done, either make sure your anti-virus software is up to date, or get you one. This will give you a reasonnable amount of security.

3. If you followed me this far, well thank you !!!

Now, welcome to a brave new world !

Make a little firewall...

[mnordstr]

I have 2 Windows computers at home that have public static IPs. Instead of using my DSL router on the windows machines, I've given them local IP addresses (192.168.*.*) and route them through my Linux server. There I've put up an iptables firewall with DNAT and SNAT, so that when the windows computers are routed through the firewall, they get their public IPs assigned to them, and you can access the computers from the outside with the public IPs. On the Linux router I've added tons of rules, and one of the most important rule is the one that blocks ports 0-1024 on all windows machines. All important ports are usually below 1024, so I can basically run filesharing, etc. without having to worry about users accessing the files from the Internet (or accessing windows messaging). However, since all ports above 1024 are unblocked, and have a public IP due to the SNAT, the users on the windows machines can use P2P apps, play games online, etc. since their machines are accessible from the outside. This has worked extremely well for a long time, no need for firewalls on the windows boxes (like Norton Internet Security). I haven't experienced any viruses, hackers or unwanted pr0n sent to the printers because of open ports. :-)