Re: Strange Socket problem?
Esmond Pitt wrote:
and if the alligator isn't there does the read time out?
No. It is really hung up. That's why I asked about it to see if
anybody else had seen this. I had never come across a situation where
you couldn't close the socket or timeout and have the code leave the read().
I have three identical programs running in the field that this was
occurring on. They all connect to the same server hardware but from
different locations through the internet. It happened first to one then
two then all three sites and continued to happen every day for about a
week. Then one day it was only one site and now it has been a week
without any sites getting hung up.
It is really bizarre and it has been a problem because either I or the
guy I work for had to restart the programs every morning and deal with
unhappy customers. I am not unfamiliar with Java socket communications
and have never had anything like this happen before. When I got away
from the OS/2 and moved to Java I thought this kind of crazy stuff would
go away. It has for the most part but every now and then something
really strange shows up. If I could get multi-headed video to work on
some version of Linux, I would probably migrate to that. Until then I'm
at the mercy of Redmond I guess.
--
Knute Johnson
email s/nospam/knute/