On Fri, Oct 13, 2000 at 09:12:30AM +0100, Yann Golanski wrote:
> On Thu, Oct 12, 2000 at 06:07:46PM +0100, Steve Haslam wrote:
> > When I first started using Exim, I used eximon to keep an eye on what
> > it was doing. However, since then I've discontinued this because:
> >
> > 1. eximon needs to be run as root/mail on the server, and needs the X
> > libraries on the server...
> >
> > 2. eximon is fine for monitoring one machine, but 3?
>
> 3? why not try 34 machines... ;>
Indeed :}
> > was to replace the single monitor program with a client/server
>
> That does sound good. What does it monitors? How does it do it?
At present, the server monitors the log files and can copy them to
clients that request them. Although clients can specify which logs
they're interested in, so it's possible to just get paniclogs from all
machines and mainlog from 1 or 2...
The client can ask the server for a listing of the queue on a
particular machine. The server has its own queue scanner rather than
running exim -bp: this has it's good and bad points, but I think it's
the better solution. (Although it does mean that cases such as
splitting the spool directory aren't supported yet).
I've basically looked at what needs to be in the communication to get
the same as eximon does, which I've got on the monitoring
side. Performing actions on the queue is trickier, I think.
SRH
--
+ Steve Haslam | W: +44-207-447-1839 +
+ Production Engineer, Excite UK | M: +44-7775-645618 +
. NP: Kryptonite (3 Doors Down) .
