On 2013-06-09 11:39 AM, wie...@porcupine.org (Wietse Venema)
<wie...@porcupine.org (Wietse Venema)> wrote:
Charles Marcus:
Would verbose logs maybe shed more light on this?
No, that would not explain why a connection fails with resource
temporarily available, because THAT problem happens inside the
KERNEL.
What would help is increase the KERNEL system-wide (not: per process)
limits on the number of file handles, network sockets, and so on.
What would also help is running lsof and understanding what
your sockets are tied up with.
Wietse (and everyone else who tried to help) - thank you thank you thank
you!
Googling details for that last clue stick was enough to strike a chord
in my memory...
I was sure I was using the same kernel, but after some googling, I
remembered something... I *had* built a new kernel a little while back
(even labeled it NEW), and changed it to the default, but did this
*without testing it*.
<facepalm>
Well, apparently I changed *something* in it that was causing this
problem, as switching back to the original kernel solved the problem...
<whew>
Thanks again!
--
Best regards,
Charles