-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi Bernard,
many thanks for replying and clarifying that point.
The machine that I'm referring to as the 'Collector' is the one running
'gmetad', but I accept that all the other 'gmond' processes are equally
capable of collating the data in that multicast channel.
However, following your suggestion, the situation is still as follows:
the nexenta machine's 'gmond' reports plenty of data about itself:
<GANGLIA_XML VERSION="3.1.7" SOURCE="gmond">
<CLUSTER NAME="infra" LOCALTIME="1304425333" OWNER="XYZ"
LATLONG="unspecified" URL="unspecified">
<HOST NAME="nexenta" IP="10.aa.bb.cc" REPORTED="1304157547"
TN="267785" TMAX="20" DMAX="0" LOCATION="XYZ" GMOND_STARTED="1304075826">
<METRIC NAME="load_one" VAL="0.18" TYPE="float" UNITS=" "
TN="267781" TMAX="70" DMAX="0" SLOPE="both">
</METRIC>
.
.
</HOST>
</CLUSTER>
</GANGLIA_XML>
but: the other machines' 'gmond' processes report plenty of data about
themselves, yet, in their XML, the entry for the nexenta machine is
always like this:
<HOST NAME="nexenta" IP="10.aa.bb.cc" REPORTED="1304425538" TN="3"
TMAX="20" DMAX="0" LOCATION="unspecified" GMOND_STARTED="0">
</HOST>
- --which is hopeless.
So it seems that somehow the METRICs, etc for the nexenta box are not
propagating out into the multicast channel, even though that machine has
announced itself.
Once again, thanks in advance to anyone who can suggest a way forward,
Best Regards
[NG]
On 04/28/2011 08:50 PM, Bernard Li wrote:
> Hi Nick:
>
> On Wed, Apr 27, 2011 at 8:41 AM, Nick Gresham
> <n.gres...@manchester.ac.uk> wrote:
>
>> However the *multicast* packets received by the Collector from the
>> Nexenta machine seem to be incorrect or deficient:
>>
>> nc 127.0.0.1 8651 | less
>>
>> [...]
>> <HOST NAME="nxxx.xxx.xxx" IP="10.xx.yy.zz" REPORTED="1303917322"
>> TN="15" TMAX="20" DMAX="0" LOCATION="unspecified" GMOND_STARTED="0">
>> </HOST>
>> [...]
>>
>> There are never any METRICs reported for the Nexenta host, and of course
>> the graphs are broken with 'gmond_started' stuck at the beginning of the
>> epoch!
>
> Port 8651 is actually for gmetad, not gmond. You should actually try
> the following on the "Collector" host:
>
> nc localhost 8649
>
> and see if what you are seeing are the same as if you run `nc 10.xx.yy.zz
> 8649`.
>
> BTW, with Multicast setup, there isn't really a notion of a
> "collector" since all hosts running gmond should have the same metric
> information of everybody subscribed to the multicast channel.
>
> Are you running the same versions of gmetad and gmond across the
> board? Do you have any 3.0.x instances in the mix?
>
> Cheers,
>
> Bernard
- --
N.J. Gresham
Unix Systems Administrator
Faculty of Life Sciences
University of Manchester
int: 7759349
ext: 0790-9893684
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/
iEYEARECAAYFAk2/9iAACgkQoqZzfMI0UdmY1QCgjgwhXaq+LrKtLbxy8HKHsrg+
CQ4An36ppv40NeEKe+6nyWJ8Jajam+YM
=1AFw
-----END PGP SIGNATURE-----
------------------------------------------------------------------------------
WhatsUp Gold - Download Free Network Management Software
The most intuitive, comprehensive, and cost-effective network
management toolset available today. Delivers lowest initial
acquisition cost and overall TCO of any competing solution.
http://p.sf.net/sfu/whatsupgold-sd
_______________________________________________
Ganglia-general mailing list
Ganglia-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ganglia-general
