Re: Port Sensitivity Test too??

[Martin Pala]

I think the count parameter has no sense in port testing. ICMP is kind 
of special case, because it is unreliable transport protocol much like udp.

Port testing is different at least for tcp based protocols - they are 
connection oriented and the possible retransmission is managed by kernel 
automaticaly - there's no need on monit side to send more then one 
request per cycle.

The general event ratio rules which should provide you the functionality 
you requested is on the wishlist as mentioned in previous mails. Anybody 
is welcomed to implement this feature ;)

Martin


M. D. Parker wrote:
> I like the ICMP sensitivity COUNT parameter.  Could this be extended to port
> testing as well......the ONE probe in causes  many up/down alerts in rapid
> succession too (i.e. false alerts really).
>
> ==========================================
>
> M. D. Parker
> Systems Administrator
> General Atomics / Electromagnetic Systems
> +1 858 455 2877
> address@hidden
>
>
>
> --
> To unsubscribe:
> http://lists.nongnu.org/mailman/listinfo/monit-general