Re: [Ganglia-general] 2 clusters in same subnet
--- regatta [EMAIL PROTECTED] wrote: Hi everyone I have one comment about ganglia document and one question, my comment is that there is no REALLY document about how to use/ ganglia (please don't ask me to read http://ganglia.info/docs/, it's the worse document I every saw, it suppose that you are expert in ganglia) you may have a point here. why notv become an expert and write new docs ? :-) Now my question :) : I have two clusters in the same subnet (each cluster has 24 nodes) , now why they are the same subnet, this is different subject but they must be :) now how can I configure one node in each cluster to run gmetad and the php-web to display the 2 clusters as 2 clusters or grids what I did is that I installed gmond in all nodes (in both clusters), I changed /etc/gmod.conf in cluster A to : cluster { name = Cluster A } and in cluster B cluster { name = Cluster B } but when I go to gmetad I find it sometime it collect them all together or it put some node in A to be B and some B to A !! Any help ? You need to separate the ports where your clusters multicast. Default is 8649. Select another port for (8648) for your second cluster. Then you need to define two datasources in gmetad.conf (you only need one of those). data_source cluster 1 node_in_cluster_1:8649 data_source cluster 2 node_in_cluster_2:8648 That should do the trick. Martin -- Martin Knoblauch email: k n o b i AT knobisoft DOT de www: http://www.knobisoft.de
Re: [Ganglia-general] Disk I/O in the Linux gmond
I do this, currently. It goes something like this, on linux. Run from cron, once a minute: #!/bin/sh CLIENT=/usr/bin/gmetric /usr/bin/iostat -x 4 2 cciss/c0d0 | grep -v ^$ | tail -4 /tmp/disk-io.tmp UTIL=`grep cciss /tmp/disk-io.tmp | awk '{print $14}'` WRITE=`tail -1 /tmp/disk-io.tmp | awk '{print $9}'` READ=`tail -1 /tmp/disk-io.tmp | awk '{print $8}'` IOWAIT=`head -2 /tmp/disk-io.tmp | tail -1 | awk '{print $4}'` $CLIENT -t uint16 -n diskio-util -v $UTIL -u '%' $CLIENT -t uint16 -n diskio-write -v $WRITE -u 'wkB/s' $CLIENT -t uint16 -n diskio-read -v $READ -u 'rkB/s' $CLIENT -t uint16 -n diskio-iowait -v $IOWAIT -u '%' and that's all there is to it. replace cciss/c0d0 with the drive device of your choice. -john - Original Message From: Martin Knoblauch [EMAIL PROTECTED] To: [EMAIL PROTECTED]; ganglia-general@lists.sourceforge.net Sent: Mon Jan 30 10:35:02 2006 Subject: Re: [Ganglia-general] Disk I/O in the Linux gmond Richard, for Solaris we have {b,l,ph}{read,write} metrics, that would cover disk-io (not latency). Nothing has been done for the other platforms in the ganglia monitor-core. Have you looked at the repository page? http://ganglia.sourceforge.net/gmetric/ Entry #39 may be interesting for you. Martin --- [EMAIL PROTECTED] wrote: Has anyone extended the Linux gmond to include disk I/O or disk latency stats? kind regards, richard grevis -- Martin Knoblauch email: k n o b i AT knobisoft DOT de www: http://www.knobisoft.de --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://sel.as-us.falkag.net/sel?cmd=lnkkid=103432bid=230486dat=121642 ___ Ganglia-general mailing list Ganglia-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/ganglia-general
Re: [Ganglia-general] Pointers on architecting a largescale ganglia setup??
Rick Mohr wrote: The unicast approach does save on gmond memory usage as you mentioned. It's up to each site to determine just how much memory the metrics will take up, and if it is considered a significant amount. (But it can get somewhat big on a large cluster like mine with a bunch of added metrics.) Can you share any code you've written for additional metrics? Thanks. BTW: Thanks eveyone for the pointers. --joel