Mail Archives: cygwin/2008/08/22/01:00:47
> You didn't look at this list archive, where I reported this problem long
> ago, and to the syslog-ng list, where somebody reported it on Linux,
> also long ago. By long ago I mean 3 or 4 versions back, but the problem
> is still there.
Woah - simmer down cowboy.
To be fair René, as I said, I did search this list, and I did see your email from back around 2006, but "restart it a few times" isn't much of a solution. Additionlly, I wasn't sure if we had the same problem since you reported syslog-ng using 100% of your CPU, and mine only seems to need 50%, and I only have one, single core CPU. I suppose I could have been more clear about seeing your "solution" but I didn't think it was necessary. I apologize. In fact, I actually thought about emailing you or Corinna directly about this, but I didn't know if that was proper list etiquette.
> Only workaround: restart syslog-ng.
> The problem for me only happens at boot time and seldom (about once a
> month = 1/30 boots). Syslog-ng goes to 100% CPU and still works if you
> test it but no other processes can do much, and those with lowered
> priority, like exim, can't do anything.
> I've also had another server process do the same, only once, I can't
> remember which one; the point is perhaps there's something else causing
> the problem, not syslog-ng.
I have debugged a few other problems recently for various cygwin services--their problems all seemed to be related to which different environment variables are set whether you log in on the windows console and then run the cygwin application vs. when one logs into cygwin via ssh. Matters are also complicated when dealing with all the different windows 2000 permission models.
If I find a solution to this syslog-ng problem I'll let the list know.
thank you,
Grav
--
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Problem reports: http://cygwin.com/problems.html
Documentation: http://cygwin.com/docs.html
FAQ: http://cygwin.com/faq/
- Raw text -