From owner-freebsd-hackers  Tue Jan  9 15:23:15 1996
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id PAA17527
          for hackers-outgoing; Tue, 9 Jan 1996 15:23:15 -0800 (PST)
Received: from cls.net (freeside.cls.de [192.129.50.1])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id PAA17519
          for <hackers@freebsd.org>; Tue, 9 Jan 1996 15:23:11 -0800 (PST)
Received: by mail.cls.net (Smail3.1.29.1)
	  from allegro.lemis.de (192.109.197.134) with smtp
	  id <m0tZm6b-0011lMC@cls.net>; Tue, 9 Jan 96 22:01 GMT
From: grog@lemis.de (Greg Lehey)
Organisation: LEMIS, Schellnhausen 2, 36325 Feldatal, Germany
Phone: +49-6637-919123
Fax:   +49-6637-919122
Reply-To: grog@lemis.de (Greg Lehey)
Received: (grog@localhost) by allegro.lemis.de (8.6.9/8.6.9) 
	id WAA23824; Tue, 9 Jan 1996 22:48:11 +0100
Message-Id: <199601092148.WAA23824@allegro.lemis.de>
Subject: Re: Using `ping' to diagnose network connections reasonable?
To: jkh@time.cdrom.com (Jordan K. Hubbard)
Date: Tue, 9 Jan 1996 22:48:10 +0100 (MET)
Cc: hackers@freebsd.org (FreeBSD Hackers)
In-Reply-To: <24751.821114877@time.cdrom.com> from "Jordan K. Hubbard" at Jan 8, 96 07:27:57 am
X-Mailer: ELM [version 2.4 PL23]
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
Sender: owner-hackers@freebsd.org
Precedence: bulk

Jordan K. Hubbard writes:
>
> I'd like to add some code to sysinstall which will attempt to
> `diagnose' a link before accepting the configuration parameters,
> catching a lot of adapter misconfiguration and incorrect data errors
> that sysinstall misses now (to fail less gracefully later).  My
> question is whether or not `ping' is a reasonable way to measure
> connectivity between your host and the gateway & dns machines.  Is it
> reasonable to assume that if a host supports forwarding or DNS
> queries, it will also answer pings?  What if you've got pings blocked
> somehow but allow DNS traffic through?  I wouldn't want to flag a host
> as `unreachable' when in fact it would have worked fine for its
> intended purpose!  That would be worse than no error checking at all.

Well, here are three possibilities to consider:

1.  I have an ISDN dialup connection, and I don't like people costing
    me money by pinging me, so I have got my ISP to drop ICMP packets
    at his end.

2.  I have an PPP dialup connection, and I haven't got my ISP to drop
    ICMP packets at his end.  Still, setting up the connection takes
    so long that the first 20 packets fall into the bit bucket.

3.  I have a machine with an independent IP processor.  If the main
    machine fails, you can't talk to it (how about that), but you can
    still ping it.  This isn't made up, I really do have a machine
    like that (based on a 68020, would you believe, but the IP
    processor is an 80386 :-).

In each of these cases, your ping check may come to an incorrect
conclusion.  Pinging is nice for an approximation, but it's no
substitute for the real thing.

Greg