Re: Able to detect disconnect, except one weird case...
- From: eselk2003@xxxxxxxxx
- Date: Mon, 5 May 2008 11:19:40 -0700 (PDT)
On May 5, 8:46 am, "Volodymyr M. Shcherbyna"
<v_scherb...@xxxxxxxxxxxxxxx> wrote:
Just a hint, enable SO_KEEPALIVE option on a socket
(http://msdn.microsoft.com/en-us/library/ms740476.aspx) and see if it helps.
I imagine it would, although I would need to change the default times
(using SIO_KEEPALIVE_VALS) to a much smaller value (the one that
defaults to 2 hours)... since I'm seeing this happen after less than
30 minutes (could be as little as 5 minutes).
I guess just the fact that so many people are using this option tells
me that keep-alives are indeed needed, and no way around them... so
I'll go ahead and create my own keep-alives, since this is easy to add
to my design (and I don't like using functions where the help file
points to a standards commitee if I can avoid it easily).
.
- Follow-Ups:
- Re: Able to detect disconnect, except one weird case...
- From: Volodymyr M. Shcherbyna
- Re: Able to detect disconnect, except one weird case...
- References:
- Able to detect disconnect, except one weird case...
- From: eselk2003
- Re: Able to detect disconnect, except one weird case...
- From: Volodymyr M. Shcherbyna
- Able to detect disconnect, except one weird case...
- Prev by Date: Re: Able to detect disconnect, except one weird case...
- Next by Date: select() always blocks non-blocking socket
- Previous by thread: Re: Able to detect disconnect, except one weird case...
- Next by thread: Re: Able to detect disconnect, except one weird case...
- Index(es):