Hello,
Logging to a pipe is dangerous, because if no one reads from pipe write system
call will block after internal kernel buffer is full.
If write is not non-blocking and is not wrapped into
poll/select/epoll/whatever else (Normally, log calls are blocking), the whole
dnsmasq process will
Oooops, logging to a pipe is not supported. Got the point ...
Sincerely,
Joachim
___
Dnsmasq-discuss mailing list
Dnsmasq-discuss@lists.thekelleys.org.uk
http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
On 13.06.2016 18:04, Eric Hiller wrote:
"This paper describes an approach with rsyslogd, an alternative
enhanced syslog daemon natively supporting MySQL and PostgreSQL. "
http://www.rsyslog.com/doc/v8-stable/tutorials/database.html
Hi.
Unfortunately this does not fit my needs. Since my
Depending upon your distro, you may be running rsyslogd anyways. Debian
defaults to it at least, I think red hats do too, but I am not sure on that.
-Eric
On Mon, Jun 13, 2016 at 11:37 AM Joachim Zobel
wrote:
>
> On 13.06.2016 18:04, Eric Hiller wrote:
> >
> > "This
Albert,
That is a great idea! I used to do just this, worked wonders. Joachim,
see the tutorial:
"This paper describes an approach with rsyslogd, an alternative enhanced
syslog daemon natively supporting MySQL and PostgreSQL. "
http://www.rsyslog.com/doc/v8-stable/tutorials/database.html
Hi
Hi Joachim,
Le Sun, 12 Jun 2016 17:39:12 +0200
Joachim Zobel a écrit:
> Hi.
>
> The main problem for my reverse_replace script is speed. It takes a
> minute, which is too slow to be run from a web gui. This is because
> the script parses the last 15k lines and puts
Hi.
The main problem for my reverse_replace script is speed. It takes a
minute, which is too slow to be run from a web gui. This is because the
script parses the last 15k lines and puts them into an IP->name lookup
tables made up from enviroment variables.
Is there a way to log queries to a