On 03/27/2014 10:07 PM, Daniel Pocock wrote: > > > I made up a rough diagram about how Ganglia 4.x could look: > > https://raw.githubusercontent.com/ganglia/monitor-core/master/doc/planning/ganglia-4.x.png > > The biggest change is the introduction of MongoDB > > Instead of having the gmetad serve up an XML every time somebody asks to > see the web page, the gmetad will just store current values into MongoDB. Well, i will post again my ideas/problems with the hope that somebody more proficient in coding and with a lot more time have the same needs as i do.
So (the big ones): 1. one of my biggest problems with gmond is that the cluster tagging is done at the level of xml and is not a per-host tag. this will allow to aggregate info from different nodes and grouping per cluster tag be done at storage level (gmetad) 2. for the storage back-end i would recommend postgres : we (our experiment) use a tool that write all the info in postgres (many hundredth thousands parameters/s) and the admins and developers assured me that postgres was the best choice. Some minor things: 1. It would be useful to have in mod_cpu all the info (or at least the 4-5 lines) of /proc/cpuinfo namely : processor, vendor_id, cpu family, model, model name (most important), stepping 2. instead of : Operating System Linux Operating System Release 3.13.0-1.el6.elrepo.x86_64 (info obtained from uname i imagine) it should print info obtained from lsb_release and uname : uname -o lsb_release -irc uname -r uname -m 3. if its possible (at least with the title of experiment) to add host_uuid variable to host variables. I understood from Daniel that something like this is already in a branch.... Thank you! Adrian
smime.p7s
Description: S/MIME Cryptographic Signature
------------------------------------------------------------------------------
_______________________________________________ Ganglia-general mailing list Ganglia-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/ganglia-general
