[smokeping-users] Smokeping/FPing scalability guidelines?
ged at jubileegroup.co.uk
Thu Jun 19 18:22:35 CEST 2008
On Thu, 19 Jun 2008, Tobias Oetiker wrote:
> Today G.W. Haywood wrote:
> > On Thu, 19 Jun 2008, Allan Eising wrote:
> > > ... 20 FPing processes ... 1997 nodes to be monitored ... the FPing
> > > processes take too long ... any guidelines for a recommended number
> > > of hosts per FPing process, so I might be able to write this into my
> > > system, to maintain scalability.
> > ... the processes running on my master smokeping host, about 1500MBytes
> > of RAM (25 percent of the memory in the host) is used by approximately
> > thirty processes (mostly FPing) ...
> not sure about these numbers you have there, you say your smokeping
> perl collector process is takeing up 1.5GB virtual memory, and that
> it is increasing over time ? This would be some sort of leak then.
No, not increasing over time. Just sitting there. A leak wouldn't be
a big deal anyway, I could just run a restart from cron. Yes, the
processes are using almost 1500Mbyte of RAM. Two-thirds of that could
be swapped if the machine needed to swap, but it doesn't. For the sake
of clarity, this isn't hurting me - it was just a response to the OP.
> if you want to code, how about improving on fping and makeing it
> scale better ?
I didn't say I want to code, but sometimes it happens that way. :)
I don't quite know what the best architecture would be, but my guess
is that it would be to have a single process which does the pinging
working with a list of hosts to ping. This feature is present in fping
already of course. As I said, I imagine that the existing tools could
do what's needed, so the first step should be to investigate what could
be done with what's already available rather than to start coding.
OTOH I was wondering what would happen if I changed the first line of
More information about the smokeping-users