Hi Steve: Care to try version 3.0.4 just to rule out this is an issue that creeped into the 3.0.5 release?
P.S. What distro/arch are you using? Thanks, Bernard On 10/17/07, Steve Okay <[EMAIL PROTECTED]> wrote: > > On Oct 16, 2007, at 13:05 , Richard Mohr wrote: > > > On Tue, 2007-10-16 at 11:27 -0700, Steve Okay wrote: > >> I'm having some problems getting gmetad to create all the rrd > >> databases with Ganglia 3.0.5. > >> > >> I'm getting: > >> ls -l /var/lib/ganglia/rrds/TSMaster/__SummaryInfo__/ > >> total 336 > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 cpu_idle.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 cpu_nice.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 cpu_num.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 cpu_percent_0.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 cpu_percent_1.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 cpu_speed.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 cpu_system.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 cpu_user.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 load_fifteen.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 load_five.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 load_one.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 proc_run.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 proc_total.rrd > >> -rw-r--r-- 1 nobody root 23640 Oct 16 12:11 uptime.rrd > >> > >> The rrds for memory usage and network are not being created despite > >> the fact that these are showing up in the XML dump I get from > >> running: > >> telnet $CLIENTHOST 8649. > > > > Is the __SummaryInfo__ directory the only one that has missing > > databases? I ran into a situation one time where all but a couple > > of my > > hosts were reporting a specific metric. However, since a few hosts > > did > > not report the metric, gmetad would not create a RRD file in > > __SummaryInfo__ for it. It was all-or-nothing. > > No, they're missing for all the nodes in the cluster. > > > > > I would first check to make sure that every host in the cluster is > > reporting the memory usage and network metrics. You may also want to > > try running gmetad in debug mode and see what it reports. > > They're all reporting back stats if you telnet to 8649: > <METRIC NAME="pkts_in" VAL="95695" UNITS="packets" TYPE="ulong" > SOURCE="beostat"/> > <METRIC NAME="pkts_out" VAL="190432" UNITS="packets" TYPE="ulong" > SOURCE="beostat"/> > > I've stopped gmetad, done an rm -rf on /var/lib/ganglia/rrds/* and > then restarted it to force it to recreate it > and it only creates the above mentioned RRDs. I've been unable to > find where in the config files or > the Ganglia code it determines what RRDs get created. I would > actually feel better in some ways if it > didn't create ANY of them because the problem would likely be much > more obvious. > > ----Steve > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Ganglia-general mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/ganglia-general > ------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ Ganglia-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/ganglia-general
