From: Harry Ohlsen <harryo@...>
Date: 2001-11-27T23:08:34+09:00
Subject: [ruby-talk:26656] Re: TCP hang problem

In article <m2wv0cbanp.fsf@zip.local.thomases.com>, "Dave Thomas"
<Dave@pragmaticprogrammer.com> wrote:

> Sounds like it might be DNS related: a DNS lookup hangs all threads.
> 
> Depending on your application, setting
> 
>           Socket.do_not_reverse_lookup = true
> 
> might help. Otherwise, you might be able to get in to the hung process
> with gdb and do a backtrace to find out what's going on.

The do_not_reverse_lookup didn't help.  I'm going to check out those
patches that were mentioned in the previous response.  If that doesn't
fix it, I'll try GDB.

Thanks Dave.

By the way, in case it gives any of you a clue, the problem was
exacerbated by moving to a newer snapshot.  With 1.6.5 stable it happened
after a few requests.  Now it happens (1.7.2 2001-11-25) on the first
write.