Re: Solution to broken Net::Ping

I am guessing its the module

guessing virtually guarantees it is not the module

Ah yes, and you are the first person to have noticed this bug since 1994. Sure.

The Net::Ping description says

You may choose one of six different protocols to use for the ping. The "tcp" protocol is the default. Note that a live remote host may still fail to be pingable by one or more of these protocols. For example, www.microsoft.com is generally alive but not "icmp" pingable.


$ perl -MNet::Ping -le " print !! Net::Ping->new->ping( shift ) " loca
+lhost
1

$ perl -MNet::Ping -le " print !! Net::Ping->new->ping( shift ) " goog
+le.com


$ perl -MNet::Ping -le " print !! Net::Ping->new( qw/syn/ )->ping( shi
+ft ) " google.com
1
[download]

A Poor Workman

Comment on Re: Solution to broken Net::Ping Download Code

Replies are listed 'Best First'.
Re^2: Solution to broken Net::Ping by solignis (Initiate) on Dec 16, 2011 at 09:16 UTC
I am aware that certain hosts are not always pingable. After a little RTFM it works now (what a surprise). Based on the above it looks like I need to create 2 ping tests (1 for icmp, 1 for tcp port test). Would it be cleaner to create the first object then undef it and recreate the object with the new options? or just create 2 different objects and destroy them after their very short lifespan? Thank you for your help	[reply]
Re^3: Solution to broken Net::Ping by Anonymous Monk on Dec 16, 2011 at 09:29 UTC
Since I don't see an official way to change protocol choice after the fact, I recommend limiting scope with functions, however many you need , maybe `sub MahPing { my ( $host ) = @_; for my $proto ( qw/ tcp icmp syn / ){ return !!1 if Net::Ping->new( $proto, 5 ) ->ping( $host ); } return !!0; }` [download]	[reply] [d/l]
Re^3: Solution to broken Net::Ping by thargas (Deacon) on Dec 16, 2011 at 14:33 UTC
If you're concerned about performance, Benchmark it, not against some production host. What works for me under my current environment, may not be the best way for you under your constraints. A general solution to monitoring availability of hosts is non-trivial. And make sure you're not optimizing prematurely. The last time I was mucking about with ping, it was on a large scale internal network and I ended up using SNMP gets instead. If you usage needs to be efficient, you might want to look at Net::SNMP as well.	[reply]