Why port blocking is bad

[torkell]

Because this is what you get if Java gets given port 3130 by Windows:



And in Eclipse I was greated by the wonderful message of:

FATAL ERROR in native method: No transports initialized
err:: No such file or directory
Error [2] in connect() call!
Socket transport failed to init.
Transport dt_socket failed to initialize, rc = -1.

Thank you, Norton.

Comments

Re: Worms?

[torkell]

No, this was just a false positive by Norton Internet Security. Norton comes configured with a list of "well-known trojan ports" which it blocks.

What happened here was that when the Java runtime (javaw.exe) asked for a free TCP port, Windows gave it one that was on Norton's block list. and so Norton blocked it, which made Eclipse (a Java development enviroment) fall over when it tried to start the debugger for my program. No trojans involved, just a trigger-happy firewall.

As to your problem, it's not something that I've seen before. The first number is the KB code for MS05-024, a patch for a remote code execution in web view. A Google for the whole file turned up nothing, but when I replaced the punctuation with spaces I got a reference on a forum to Software Update Services (specifically, about SUS going on strike and spewing out errors instead of patches). If you're running SUS then it may be worth checking the settings and relevant logs.

Apart from that I don't have any ideas. Hope you have more success than me at solving it!

(x-posted to supplied e-mail and in reply to this comment on this post)

Re: Worms?

[ricold.livejournal.com]

"just a trigger-happy firewall."

Quote of the day!