Hey Chris:
I assume that you are using different mcast ports for the different data
sources.
Typically I find that if there are some errors with the metrics after I
have done some modifications to the cluster (add/remove nodes) - I have
to stop all the daemons (gmond/gmetad) and restart them. It usually
takes some time before the data gets properly propagated and eventually
the correct information shows up.
Hope this helps.
Cheers,
Bernard
Chris Jones wrote:
I've got a set of three clusters (one with 3 computers, one with 7, one
with 23) which I'm monitoring using Ganglia. While the summary
information shown in the grid view is correct for the first two
clusters, the number of hosts and CPUs shown for the third is wildly off
(typically only a handful of hosts are listed as being up). However,
when I take a look at the cluster view, the information for the third
cluster is shown correctly.
The only thing different about this cluster is that because our routers
aren't yet correctly set up to pass IGMP and multicast packets, the
cluster is effectively partitioned into sub-clusters for gmond
purposes. These sub-clusters vary in size between one and five hosts.
To handle this partitioning, I've got multiple data sources (one for
each sub-cluster) set up in the web host's gmetad.conf. Would this be
causing the incorrect values in the grid view? If so, why isn't it
affecting the cluster view?
Cheers,
Chris
-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more productive? Does it
help you create better code? SHARE THE LOVE, and help us help
YOU! Click Here: http://sourceforge.net/donate/
_______________________________________________
Ganglia-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/ganglia-general
