Thats what I am waiting for too. I had some problem compiling the latest snapshot of ganglia-2.6.0 on Matt's website.
Couple of solutions were suggested in this thread, but none worked for me. https://sourceforge.net/mailarchive/forum.php?thread_id=4578097&forum_id=7186 Thanks for your attention. Regards, Vineet. Quoting "Adesanya, Adeyemi" <[EMAIL PROTECTED]>: > > Hi Vineet. > > I'm not a big fan of multicasts either. Right now I generate ganglia metrics > on around 150 nodes. Gmetric is also used to produce a large number of custom > metrics in addition to gmond's built in metrics. Multicast packets are sent > to every single host machine on a router/subnet. Multicast channels allow you > to label the packets but it doesn't result in any kind of network filtering. > The solution is UDP unicasts and I am eagerly waiting Ganglia 2.6.0 which > implements this feature which for me is essential if I am to stand a chance > of convincing our systems people that Ganglia will scale. Please Matt, roll > out the unicast feature. > > ---- > Yemi > > > -----Original Message----- > > From: [EMAIL PROTECTED] > > [mailto:[EMAIL PROTECTED] On > > Behalf Of Vineet Mokashi > > Sent: Tuesday, September 07, 2004 2:39 PM > > To: [email protected] > > Subject: [Ganglia-general] Multicast issue: High bandwidth usage > > > > This issue has been addressed earlier, but no fool-proof > > solution was suggested. > > Now that I have been monitoring our 32-node Beowulf, running > > RHEL-AS with 2.4.20-20.9smp kernel, the problem has started > > to look grave. > > > > Upto 60% of the packets on each node can be attributed to > > gmond and the inter-process communication is getting affected > > due to this. Turning on 'deaf' > > mode does not help, since gmetad is unable connect to 8649 to > > get any data from the gmond daemons. > > > > As a probably workaround, I (painfully) configured gmond with > > a separate multicast channel for each node, say > > 239.2.11.40-72, and removed all trusted_host entries (but for > > the login node, which runs gmetad). > > > > This solves a part of the problem - every node's gmond > > communicating with every other node to probe data. The > > bandwidth consumption has decreased significantly. The broken > > link with this set-up is that gmetad is unable to query data > > from any node except the one running on the default > > multi-cast channel. All the gmond daemons use the default XML > > port : 8649. > > > > Has anyone tried something like this before? How can we get > > gmetad to query multiple multicast channels? On top of all > > this, the latest snapshot of ganglia-2.6.0 fails to build. > > > > Regards, > > Vineet > > (Newbie Linux Administrator) > > > > ---------------------------------------------------------------- > > This message was sent using IMP, the Internet Messaging Program. > > > > > > ------------------------------------------------------- > > This SF.Net email is sponsored by BEA Weblogic Workshop FREE > > Java Enterprise J2EE developer tools! > > Get your free copy of BEA WebLogic Workshop 8.1 today. > > http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click > > _______________________________________________ > > Ganglia-general mailing list > > [email protected] > > https://lists.sourceforge.net/lists/listinfo/ganglia-general > > > ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.
