It looks like the delays are coming from the usleep line in send_ping();
this looks like it was put in there to do some network/cpu throttling,
which would definitely make the bbtest-net duration longer, but I'm not
sure yet why it is affecting the icmp return times.
Running hobbitping with the flag "--max-pps=10000000" seems to fix
things.
e
-----Original Message-----
From: Schwimmer, Eric E *HS [mailto:user-1e1008b069d5@xymon.invalid]
Sent: Monday, August 28, 2006 9:49 AM
To: user-ae9b8668bcde@xymon.invalid
Subject: RE: [hobbit] Hobbitping/fping behaviour
I have the same problem after upgrading to 4.2.0, but I reverted to the
version of hobbitping that was beta tested under 4.1.2p1, and both my
ping times and bbtest-net durations dropped considerably (1522 hosts
pinged in 17 seconds with the old hobbitping, down from 55 seconds with
the new hobbitping).
A diff against the two versions reveals quite a few changes that were
made;
I have been extremely pleased with the first version so I'll stick with
that for a while longer. I'll try running the new version through
gprof, but I am not as skilled at these things as Henrik :)
e
-----Original Message-----
From: Rob MacGregor [mailto:user-07c9d92ae079@xymon.invalid]
Sent: Friday, August 25, 2006 2:26 AM
To: user-ae9b8668bcde@xymon.invalid
Subject: Re: [hobbit] Hobbitping/fping behaviour
On 8/24/06, GALLIOT Guillaume <user-b013132bbbef@xymon.invalid> wrote:
Hello,
I have updated my hobbit installation from 4.1.2 to 4.2.0
The new hobbitping gives different results in the web interface (about
40ms
latency on a local network) than fping (0.15 ms),i tried to reproduce
these
results with ping,fping & hobbitping but they give me same results on
command line (0.15ms)
fping -e msrv999lno001
msrv999lno001 is alive (0.16 ms)
PING msrv999lno001 (10.0.4.173) 56(84) bytes of data.
64 bytes from msrv999lno001 (10.0.4.173): icmp_seq=0 ttl=64 time=0.126
ms
64 bytes from msrv999lno001 (10.0.4.173): icmp_seq=1 ttl=64 time=0.113
ms
/usr/lib/hobbit/server/bin/hobbitping 10.0.4.173
10.0.4.173 is alive (0.18 ms)
In the web interface I get 44 ms
I reverted to fping and results are now the same on command line and
web
interface.
Does anyone has the same issue ?
Yes, I saw exactly the same - hobbitping used by hobbit had a delay of
tens of ms, used on the command line had a delay of fractions of an
ms.
--
Please keep list traffic on the list.
Rob MacGregor
Whoever fights monsters should see to it that in the process he
doesn't become a monster. Friedrich Nietzsche